article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
long ago aharonov and bohm @xcite proposed `` en passant '' an extension of the electromagnetic lagrangian , obtained by adding a term @xmath6 to the minimal gauge - invariant lagrangian which leads to maxwell equations .
a previous attempt in this direction had also been made by ohmura @xcite .
this modified electromagnetic theory has been recently re - analysed by van vlaenderen and waser @xcite , hively and giakos @xcite . in @xcite
an hamiltonian formalism is employed , in @xcite a quaternion formalism , in @xcite a standard three - dimensional vector formalism . in this work
we obtain new insights into this extended theory using a standard four - vectors formalism , which is in our opinion more effective , and allows to find some useful explicit expressions missing from the other approaches .
we obtain a simple , consistent generalization of the maxwell equations which is independent from the coupling @xmath7 and has the important property of charge conservation censorship : although the equations are compatible with sources which do not respect local charge conservation , the measurable field tensor @xmath8 which solves them is always coupled to a conserved charge
. the results can be relevant for applications to condensed - matter systems with coherent quantum tunnelling , or to more exotic situations in which the occurrence of a chiral anomaly or a spacetime singularity causes a tunnelling process between two vacuum states @xcite .
the modified theory of aharonov - bohm has an additional degree of freedom , the scalar field @xmath9 , which can not be set to zero through a suitable choice of the gauge , as usually done in the standard theory .
the field @xmath1 amounts to the presence , in wave solutions , of an anomalous longitudinal component . since , however , the source of @xmath1 is the quantity @xmath10 , in fact @xmath1 is decoupled from matter if charge is locally conserved everywhere .
the authors of @xcite discuss the possibility that charge fluctuations generate fluctuations of @xmath1 , or that boundary condition at the interface between vacuum and some media may also affect @xmath1 .
@xcite briefly mentions the possibility that the local conservation of charge is violated in macroscopic tunnelling phenomena .
we shall comment on this latter point in our conclusions . in sect .
[ sec2 ] we show that working in the standard four - vector formalism it is possible to eliminate completely the field @xmath1 from the field equations , and so obtain generalized , self - consistent maxwell equations with very interesting properties , namely : ( a ) they allow to compute @xmath11 starting from any physical source @xmath12 , either locally conserved or not ; ( b ) if @xmath13 , the equations reduce to maxwell equations ; ( c ) if @xmath14 , the equations give an additional current @xmath4 such that @xmath15 is conserved and equal to @xmath16 .
it follows that any anomalous local non - conservation of the source @xmath12 is automatically `` censored '' by the field that it generates , because when we observe the source by measuring @xmath11 with test particles we can only see the conserved current @xmath15 . in spite of the censorship ,
if the physical current @xmath12 is not locally conserved , the consequences are real , because the @xmath11 field will contain a non - maxwellian component whose features will be analysed , in two simple cases , in sect . [ sec3 ] .
[ disc ] contains our conclusions and a brief discussion of possible applications to josephson tunnelling in high - t@xmath17 superconductors .
the maxwell equations in heaviside units are written in covariant form as @xcite @xmath18 note that since @xmath19 is antisymmetric , the field equation for @xmath19 can only be solved if @xmath20 .
this field equation can be derived from the lagrangian @xmath21\ ] ] @xmath22 is seen as a function of the fundamental dynamical variable @xmath23 and is the most general possible relativistic invariant lagrangian constructed with a four - vector field .
the term proportional to @xmath24 can be written as a four - divergence and therefore gives in fact no contribution to the action @xmath25 . when @xmath26 one speaks of the `` minimal '' lagrangian .
in a gauge transformation @xmath27 the variation of @xmath22 is @xmath28 if the current is conserved , this is equivalent to @xmath29 and therefore @xmath22 is gauge - invariant up to a four - divergence .
aharonov and bohm proposed to add a term @xmath30 to the minimal @xmath22 .
the new addition , not being a four - divergence , changes the field equations , as follows : @xmath31 under a gauge transformation , the variation of the aharonov - bohm lagrangian is @xmath32\ ] ] thus the lagrangian is not gauge - invariant anymore .
it is only invariant ( if @xmath33 ) under reduced gauge transformations , such that @xmath34 . in the aharonov - bohm theory
we can not impose the familiar lorenz gauge , choosing a new four - potential @xmath35 such that @xmath36 .
therefore we must regard the quantity @xmath37 as a non - trivial dynamical variable .
now take the derivative @xmath38 of the field eq .
( [ field1 ] ) and remember that @xmath39 is antisymmetric
. we obtain @xmath40 where @xmath41 denotes the dalembert operator @xmath42 . from ( [ field2 ] )
we obtain @xmath43 the well - known operator @xmath44 is linear and non - local , as can be seen passing to four - momentum space , where it is represented by @xmath45 .
( [ field3 ] ) allows us to write an expression for the variable @xmath1 , essentially integrating over the `` source '' @xmath46 : @xmath47 now go back to ( [ field2 ] ) , rename the summation index ( @xmath48 ) , take again the derivative @xmath49 and multiply by @xmath7 .
we obtain @xmath50 plug this into ( [ field1 ] ) and re - write the inhomogeneous generalized maxwell equations as follows ( the homogeneous equations remain the same ) : * version 1 * : equations for the field strength @xmath51 , the most simple and useful : @xmath52 * version 2 * : equations for the four - potential @xmath53 , formally more complex : @xmath54 note that eq .
( [ mme2]-c ) is technically an awkward gauge condition , `` obtainable '' with the residual gauge freedom .
a quantum field theory of the four - potential in this gauge looks difficult !
yet we believe the classical field equations of version 1 are interesting enough .
note that although these are derived from the aharonov - bohm lagrangian @xmath55 , the parameter @xmath7 has disappeared .
if the current @xmath56 is conserved , then the usual maxwell equations are recovered . the new current component @xmath57 which now contributes , together with @xmath58 , to generate the field @xmath39 , is such that the total current @xmath59 is always conserved , as can be checked in two ways : ( 1 ) by taking the derivative @xmath38 in the first equation of ( [ mme1 ] ) ; ( 2 ) by taking the derivative @xmath38 in eq .
( [ mme1]-b ) , which yields consistently @xmath60 summarizing , we can say that the input of the generalized electrodynamic equations ( [ mme1 ] ) is a four - current @xmath58 which is not necessarily conserved ( computed , for instance , from an `` anomalous '' microscopic model , as discussed in the following ) ; but the output is an electromagnetic field tensor @xmath39 which has the usual properties , including that of being generated by a conserved current , namely @xmath59 .
it follows the important property that at the macroscopic level the current is always conserved , as far as it is possible to measure it through the field it generates . since eq .
( [ mme1]-a ) is linear , the field @xmath39 is the sum of the fields generated by the currents @xmath58 and @xmath57 . in general , the difference between the two currents is that even if the `` primary '' current @xmath3 is confined in a certain region of spacetime , the `` secundary '' current @xmath57 is not , because of the non - local expression which relates it to @xmath58 .
a further discussion of this point is given in sect .
let us find the solutions of the generalized maxwell equations ( [ mme1 ] ) for a simple planar source in stationary flow regime @xmath61 ( fig . 1 ) .
we assume that inside the source the three - divergence @xmath62 is not zero , i.e. , charge is being emitted ( possibly by tunnelling from another region where it is absorbed ; compare fig . 2 ) .
the source is a thin layer of thickness @xmath63 on the @xmath64-@xmath65 plane .
suppose that inside the layer we have @xmath66 ( fig . 1 , ( a ) ) . the current density @xmath67 will generate , according to eq .
( [ mme1]-a ) , a magnetostatic field in the space region with @xmath68 .
there will also be , however , an additional magnetostatic field generated by the secondary current @xmath69 . for a plane which emits a stationary flow of charge , thus violating the continuity condition @xmath70 . the secondary current density ( eq .
( [ mme1]-b ) ) is @xmath71 ( + for @xmath72 , for @xmath68 ) , so the violation is censored by the total current @xmath73 .
compare fig . 2 , where the secondary currents are explicitly drawn.,width=377,height=264 ]
let us compute @xmath69 .
start from eq .
( [ mme1]-b ) and rewrite it as @xmath74 . in this case @xmath75 does not depend on time .
then @xmath76 and for the spatial component @xmath77
we can write @xmath78 but @xmath79 is , in heaviside units , like @xmath80 the electric potential of a charge distribution with density @xmath75 ; and @xmath81 is like the electric field of this distribution . these are mathematical properties familiar from the maxwell equations , which we may apply here , although in a different context .
the source of this `` equivalent electric field '' , which coincides mathematically with the secondary current @xmath69 , is a planar charge distribution with surface density @xmath82 therefore the secondary current density @xmath69 enters into the @xmath64-@xmath65 plane both from the right and the left ( like the electric field of a negatively charged plane ) :
@xmath83 it follows that the total current density is continuum : @xmath84 the emission of charge from the plane is censored , since an observer who can only measure the electromagnetic field will see a continuum flux of charge going from @xmath85 to @xmath86 . for the case of a double plane ,
. 2 . , which amounts to `` four units '' everywhere .
@xmath87 : physical , or `` primary '' , current density .
@xmath88 : secondary current density emitted by plane 1 according to eq .
( [ mme1]-b ) .
@xmath89 : secondary current density emitted by plane 2.,width=377,height=264 ] next consider a stationary dipolar source made of a pointlike charge emitter placed near a pointlike absorber , i.e.@xmath90\ ] ] let us just compute the secondary current , in order to check that it is not localized : @xmath91 = \frac{1}{2}\mu { \partial ^k}\left ( { \frac{1}{{\left| { { \bf{x } } - { \bf{a } } } \right| } } - \frac{1}{{\left| { { \bf{x } } + { \bf{a } } } \right| } } } \right ) \label{i - dip}\ ] ] therefore @xmath77 is like the electric field of a dipole . the magnetostatic field that it generates can be computed from ( [ mme1]-a ) , for instance through the biot - savart formula .
we have seen that the generalized aharonov - bohm electrodynamics , here re - formulated and solved for the scalar field @xmath1 , is applicable also to situations where charge is not locally conserved . in these situations
the electromagnetic field tensor @xmath8 hides the non - conservation , because its source contains also an additional , `` secondary '' current term which compensates for the non - conserved part of the physical current .
this mechanism is particularly clear in the example of the double plane with tunnelling ( fig . 2 ) .
one might thus have the feeling that in the end the theory is not telling us anything new , compared to maxwell theory .
this is not the case , however , since the field generated by the secondary current @xmath4 is not maxwellian , as a consequence of the non - local character of @xmath4 .
such non - locality descends in general from eq .
( [ mme1]-b ) and is clearly seen in the example of the static dipole , eq .
( [ i - dip ] ) .
( an oscillating dipole solution will be presented in a separate paper . )
what do we mean exactly by `` non - maxwellian '' field ?
given any antisymmetric tensor @xmath5 , according to ( [ mme1]-a ) the derivative @xmath2 yields a conserved source @xmath15 which generates it .
if we substitute this entire source into the maxwell equations and regard it as a physical source , then we obtain the same @xmath5 as a solution of the maxwell equations , and so by definition `` maxwellian '' .
it may happen , however , that @xmath15 is manifestly not physical , for instance because it extends to spatial infinity while the material systems we are considering are limited in space . in this case
we say that @xmath5 is not maxwellian .
more specifically , we may wonder if there exists a solution @xmath92 which represents a wave in vacuum ( @xmath93 ) with a longitudinal component . from ( [ mme1]-a )
we obtain @xmath4 : @xmath94 .
( [ mme1]-b ) relates @xmath4 to the physical current density @xmath3 .
the question then becomes : does a suitable physical current @xmath3 exist , located inside a bounded region , able to generate in the vacuum space near it a current @xmath95 ?
a detailed answer to this question will require to analyse the invertibility of eq .
( [ mme1]-b ) .
the answer might be affirmative thanks to the non - locality of the operator @xmath96 .
compare also the functional approach in @xcite .
finally , let us briefly discuss the possible concrete existence of physical currents which are not locally conserved .
the abj anomalies represent , as mentioned , such a possibility .
we recall that abj anomalies have also been observed in condensed matter @xcite , but as anomalies in momentum space , and therefore probably not relevant to our case .
the possible non - conservation of current in macroscopic quantum tunnelling is discussed in @xcite , with reference to josephson junctions in high - t@xmath17 superconductors . in short ,
it is conjectured that due to the very small correlation length @xmath97 , the stationary continuity condition @xmath98 may be violated in wide inter - grain junctions where @xmath99 decreases by many magnitude orders @xcite while @xmath100 can not increase in inverse proportion .
it is known that the solutions of the ginzburg - landau equation satisfy a continuity condition for the current .
the ginzburg - landau equation , however , should be only regarded as a phenomenological local approximation .
the microscopic theory of high - t@xmath17 superconductors is still unsettled . in complex condensed matter systems ,
macroscopic wave functions of charge carriers obey a constrained equation and have therefore in general a non - local conserved current @xcite .
this might be related to the puzzle of the superluminal tunnelling times @xcite . in conclusion , it can not be ruled out that in certain situations the macroscopic tunnelling of the wave function is associated to non - local conservation of the current , thus requiring the generalized aharonov - bohm theory , and possibly generating non - maxwellian fields in the sense discussed above . furthermore , the possible existence of non - conserved currents associated to tunnelling processes is obviously relevant also for other interactions , besides the electromagnetic interaction .
one can consider , for instance , non - conserved hadronic currents .
the generalized electrodynamics presented in this paper can be applied to such currents and tells us that they could be partially concealed from electromagnetic measurements , but still be real .
g. modanese , `` covariant formulation of aharonov - bohm electrodynamics and its application to coherent tunnelling '' , in _ unified field mechanics ii _ , proceedings of the 10th symposium in honour of j .-
vigier , portonovo , 2016 . | the aharonov - bohm electrodynamics is a generalization of maxwell theory with reduced gauge invariance .
it allows to couple the electromagnetic field to a charge which is not locally conserved , and has an additional degree of freedom , the scalar field @xmath0 , usually interpreted as a longitudinal wave component . by re - formulating the theory in a compact lagrangian formalism ,
we are able to eliminate @xmath1 explicitly from the dynamics and we obtain generalized maxwell equation with interesting properties : they give @xmath2 as the ( conserved ) sum of the ( possibly non - conserved ) physical current density @xmath3 , and a `` secondary '' current density @xmath4 which is a non - local function of @xmath3 .
this implies that any non - conservation of @xmath3 is effectively `` censored '' by the observable field @xmath5 , and yet it may have real physical consequences .
we give examples of stationary solutions which display these properties .
possible applications are to systems where local charge conservation is violated due to anomalies of the abj kind or to macroscopic quantum tunnelling with currents which do not satisfy a local continuity equation .
* keywords * : maxwell equations ; abj anomaly ; local charge conservation . |
the abundance of substructure in galaxy halos is emerging as a key test in theories of galaxy assembly . in cold
dark matter cosmogonies , dark matter overdensities collapse to form cusped halos , with the smallest and least massive halos being the densest .
the simulations of @xcite and @xcite predicted hundreds of small galactic satellite halos , in contrast to the nine then known satellites around the milky way .
@xcite had already suggested that photoionisation may lengthen the cooling times of gas in haloes with low circular speeds .
this effect suppresses the formation of satellite galaxies , but produces a large population of entirely dark satellites ( see e.g. * ? ? ?
* ; * ? ? ?
* ) . in strong lensing
, it has been known for some years that simple , smooth models of galaxy lenses usually fitted the image positions well , but the flux ratios of the images poorly . in a bold paper , @xcite argued that the flux anomalies in gravitational lens systems could be interpreted as evidence for entirely dark substructures .
they used 7 of the then available four - image lens systems to claim detection of substructure amounting to a total mass of @xmath5 of the lens galaxy mass .
however , this interpretation was challenged as flux anomalies could arise from alternative sources , such as absorption , scattering , or scintillation by the interstellar medium of the lens galaxy , or by higher order harmonics in the ellipsoidal models used to fit lens systems , or stellar microlensing ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in particular , the surface mass density in substructure as judged from simulations seems to be lower than that required by gravitational lensing , at least within the einstein radius which probes primarily the inner parts of halos ( e.g. , * ? ? ? * ) .
it is also hard to reproduce the observed statistics on cusp violations with substructure @xcite .
this argues against substructure as a primary cause of anomalous flux ratios .
nonetheless , there is good evidence in favour of substructure , and , in some cases , visible substructure can be identified .
one of the gravitational lens systems with a flux ratio anomaly is the radio - loud quadruple class b2045 + 265 discovered by @xcite .
recent deep _ hubble space telescope _ and _ keck _ imaging of this system by @xcite have revealed the presence of a tidally disrupted dwarf galaxy g2
. this may be the cause of the flux ratio anomaly , although caution is needed as modelling suggests that g2 must be very highly flattened ( @xmath6 ) .
there is also evidence for visible structure possibly a small galaxy in the radio - loud quadruple mg0414 + 0534 .
@xcite already argued that the perturbation caused by this object may account for the relatively poor agreement between the observed data on this lens and the theoretical models . the quadruple lens systems class b1608 + 656 ( see e.g. , * ? ? ?
* ) has a loose group of galaxies at the same redshift as the main lensing galaxy , a phenomenon that also occurs in the six - image radio - loud system b1359 + 154 @xcite .
it seems that visible substructure may well be responsible for some of the flux ratio anomalies .
in addition , there have been striking observational developments nearer to home .
the last two years have seen the discovery of 10 faint , new milky way satellites in data from the sloan digital sky survey ( sdss , see willman et al .
2005 , zucker et al .
2006a , b ; belokurov et al . 2006 , 2007 ) .
it seems likely that a population of ultra - faint , dwarf galaxies does surround the milky way .
these may be representatives of the `` missing satellites '' , as predicted by the simulations of @xcite and @xcite , or they may be a population of tidal dwarf galaxies or even star clusters ( see e.g. , * ? ? ?
such ultra - faint objects are only detectable nearby , and so would be to all intents and purposes dark at the typical redshifts of strong lenses .
all this suggests that it is worth re - examining the effects of substructure on strong lenses . in this paper , we answer the following questions .
given what we know about the satellite populations , how frequently might we expect anomalous flux ratios for elliptical and spiral galaxies ?
if luminous satellite galaxies represent the bright and massive end of a predominantly faint population of objects , how frequently might we expect to attribute flux ratio anomalies to visible objects ?
three ( b2045 + 265 , mg0414 + 0534 , b1608 + 656 ) out of the sample of six quadruplets originally identified by @xcite as anomalous have optically identified companions that are possible causes . at the outset
, this seems surprisingly high , if the substructure is predominantly dark .
the paper is arranged as follows . in 2 ,
models of elliptical and spiral galaxies are briefly introduced , together with their satellite galaxy populations . in 3 , the frequency with which anomalous flux ratios occur is shown to depend primarily on the scalelength of the satellite distribution , the mass model used for the satellites and the total mass in substructure , while depending only weakly on how the mass is distributed between satellites .
the simulations reported in 4 give the expected fraction of anomalous flux ratios for both ellipticals and spiral lenses , together with the typical numbers caused by high mass and luminous satellite galaxies .
for convenience , we follow @xcite by defining in the lens plane the dimensionless distance , dimensionless surface mass density and critical surface mass density @xmath7 with @xmath8 being the distances to the source , lens , and between lens and source , and @xmath9 an arbitrary scale length which relates the scaled ( uncapped ) coordinates to the unscaled ( capped ) .
the corresponding dimensionless deflection potential is the solution of the poisson equation @xmath10 and the dimensionless bending angle is @xmath11 we examine two models for the main lens galaxy .
the first is appropriate for an elliptical galaxy lens .
it is a pseudo - isothermal elliptic deflection potential ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) @xmath12 where @xmath13 is a dimensionless core radius ( the length scale being the arbitrarily chosen @xmath9 ) , and @xmath14 is the dimensionless einstein - ring radius of the singular isothermal sphere with line - of - sight velocity dispersion @xmath15 corresponding to the @xmath16 case .
there are two critical curves : a small inner one which maps to a ` radial ' caustic and an outer ` tangential ' one which maps to an astroid caustic . at the ranges of @xmath17 we consider ,
the astroid caustic is wholly within the outer caustic .
a point source has one image if it is outside both caustics , three if it is inside the outer caustic , and five if it is inside the astroid caustic .
triplets and quintuplets , however , are effectively doublets and quadruplets , because one of the multiple images is a highly demagnified central image within the small inner critical curve . for concreteness
, we consider the elliptical potential to have ( unless otherwise specified ) a velocity dispersion @xmath18 and a core radius of @xmath19 , as suggested by @xcite and @xcite .
we restrict the ellipticity parameter to be smaller than @xmath20 , otherwise the corresponding surface mass density becomes dumbbell - shaped , which is inappropriate for elliptical galaxies ( see * ? ? ? * ) .
the second model is appropriate for a spiral galaxy lens .
it is a three - component model widely used in galactic astronomy ( see e.g. @xcite , @xcite ) as a model for the milky way .
it has a @xcite bulge , a miyamoto - nagai ( 1975 ) disc and a cored isothermal halo .
the hernquist bulge has 3d mass distribution @xmath21 where @xmath22 is the spherical polar radius and @xmath23 a core radius . that of the halo is @xmath24 where @xmath25 is the core radius and @xmath26 the central density .
the mass distribution of the miyamoto - nagai disc is complicated , but the deflection potential in the edge - on case is simple : @xmath27\:,\ ] ] where @xmath28 is the dimensionless mass and @xmath29 and @xmath30 control the shape of the distribution .
we normalize our disc galaxy lens to the milky way , according to the parameters given in @xcite .
disc galaxies give rise to three main different classes of multiple - image configurations ( see e.g. @xcite ) : ` core triplets ' ( in effect , doublets ) , ` disc triplets ' ( where images straddle the plane of the disc ) and quintuplets ( in effect , quadruplets ) .
the small 7-imaging butterfly cusp in the caustic of this edge - on milky way is ignored here ( see e.g. @xcite ) .
we choose the redshift of the lens to be @xmath31 and that of the source to be @xmath32 .
these are the median redshifts of known 4-image lens systems ( see the castles website ) , omitting those known to have more than one main lens and those without known lens and source redshifts .
we use a flat @xmath33cdm cosmology with @xmath34 , @xmath35 , @xmath36 .
the plummer model is often fitted to observed dwarf spheroidal galaxies ( see e.g. @xcite , @xcite ) .
we give our satellites densities @xmath37 where @xmath38 @xmath39 and @xmath40 being cartesian coordinates with their origin at the centre of the @xmath41th satellite galaxy , @xmath42 is the central density of that galaxy , and @xmath43 where @xmath44 is the plummer model scale radius .
the mass of any satellite is @xmath45 we give a ` typical ' @xmath46 plummer satellite a scale radius ( equal to its half - light radius ) of @xmath47 which is the median half - light radius of known milky way satellite spheroidal galaxies ( see e.g. @xcite ) , and vary @xmath48 as @xmath49 . this mass - radius scaling is consistent with those found in @xmath50-body simulations , where the size of satellites scales as a power of the mass .
we also put a lower limit on @xmath48 of @xmath51 , the smallest known half - light radius of a milky - way dwarf spheroidal @xcite .
we also examine a second density profile for the satellites , motivated by numerical simulations .
the nfw density profile @xcite @xmath52 is found to be a good fit to cold dark matter subhalos . rather than using the nfw , which falls off as @xmath53 at large radii and therefore has formally infinite total mass
, we use the hernquist density profile , which is similar to the nfw profile , but has an @xmath54 asymptotic density decay .
our hernquist satellites are less concentrated than the plummers , and we make a lowest order estimate of the effects of tidal stripping by truncating them at tidal radii @xmath55 defined by the condition @xmath56 where @xmath22 is the radial distance of the @xmath41th satellite from the centre of the main galaxy , @xmath57 its mean mass density , and @xmath58 the mean mass density due to the main galaxy enclosed within a sphere of radius @xmath22 when the pseudo - isothermal elliptical potential is approximated by a singular isothermal sphere , that is , @xmath59 the truncated hernquist profile is then @xmath60 where @xmath61 is the radial distance from the centre of the @xmath41th satellite , @xmath62 are the untruncated hernquist masses and @xmath63 are hernquist scale lengths .
the scale length of each satellite is chosen so that the location of the peak circular velocity of the model ( which is @xmath64 ) varies with the mass bound within the tidal radius @xmath65 as @xmath66 in agreement with the numerical simulations of , for example , diemand , kuhlen & madau ( 2007 ) .
if the tidal radius of any hernquist satellite is less than its scale radius , we reject it as being too strongly tidally disrupted to be approximated by this mass profile .
for example , @xcite find that nfw clumps lose their mass rapidly for @xmath67 .
it is beyond the scope of this paper to consider lensing effects due to the tidally - stripped matter from dwarf galaxies .
the distribution of satellite galaxies is taken as spherically symmetric .
explicitly , we assume that the number density in the lens plane is @xmath68 where @xmath69 determines the asymptotic fall - off of the distribution , and @xmath70 is a central softening parameter .
we choose @xmath71 so that the three - dimensional number density , obtained by abel deprojection of ( [ eq:2-satellite-2d - density ] ) , falls off like @xmath72 , similar to the behaviour observed in the milky way ( see e.g. , * ? ? ?
we choose two different @xmath70 , so that half the satellites are within a ( spherical polar ) radius @xmath0 of @xmath73 or @xmath74 respectively of the centre of the main galaxy .
the distributions are plotted in fig .
[ fig : satellitepdf ] .
note that , at least as judged from the case of the milky way , we expect half the satellites to lie within @xmath75 kpc , and so @xmath76 is perhaps the more realistic .
= 6 cm for a given source position , the lens equation can be solved numerically by first triangulating the image plane , mapping the grid to the source plane to triangulate the image positions , and then using a multi - dimensional newton - raphson procedure .
the satellite galaxies have small effects on the positions of the images and on critical curves and caustics .
their effect is mainly on the ratios between image fluxes .
when a satellite galaxy is sufficiently near an image , it changes the magnification of that image .
only if the separation between the satellite and the image is still smaller does image splitting occur . in this paper , we focus on the effect of the satellites on the flux ratios of lenses rather than the image multiplicity .
when images split , they are near new critical curves created by the satellite galaxy and are highly magnified , which certainly results in a flux ratio change .
we generate positions of sources randomly to find 1000 five - image and 1000 three - image systems .
for each system , the image positions and magnifications are found numerically , both for the main galaxy alone and with satellite galaxies at positions randomly generated according to the distribution ( [ eq:2-satellite-2d - density ] ) ( or its 3d deprojection , for the purposes of finding tidal radii ) .
we count the number of systems where the satellite galaxies change the ratio of any two image fluxes by @xmath77 or more ( discounting the usually unobservable central images ) .
= 8.38 cm for the moment , let us fix the total mass in satellite galaxies @xmath78 as @xmath79 and share it equally among a varying number @xmath50 of satellites .
the lens galaxy is an @xmath80 pseudo - isothermal elliptical potential , and the satellites are spatially distributed with @xmath81 .
the results are shown in fig .
[ fig : ctm - plummer ] , which illustrates that , over a large range of @xmath82 or @xmath50 , the proportion @xmath83 of quads with flux ratios affected by the satellites is only weakly dependent on how mass is apportioned between the satellites . that is , the proportion @xmath83 is mainly sensitive to the total mass @xmath84 rather than @xmath50 or @xmath82 .
the solid line in fig . [ fig : ctm - plummer ] suggests that satellites of mass @xmath85 are only @xmath86 more efficient at altering flux ratios than those of mass @xmath46 .
the plummer scale radius @xmath48 , if set differently , can affect the results .
for example , fixing @xmath48 to @xmath87 for all @xmath82 causes @xmath88 fall off rapidly below @xmath89 , as the central density of the satellites decreases .
[ fig : ctm - plummer ] can be understood qualitatively as follows .
the magnification of an image is changed by @xmath90 and the flux ratios of the lens system affected if the image is within a ` radius of influence ' @xmath91 of a satellite . here , @xmath91 depends on the location and magnification of the image , as well as the mass of the satellite .
( @xmath91 increases with the mass of the satellite and the magnification of the image .
note that there can be an ambiguity in @xmath91 .
as a satellite moves closer to the image , the image may be first slightly brightened , before dimming rapidly as the satellite approaches , or first slightly dimmed before brightening rapidly .
so , for a large enough satellite , there may be two ranges in which the flux changes by @xmath92 .
for the purpose of our rough argument here , however , we ignore this complication . ) for a given configuration , the probability that the flux ratio is changed appreciably by satellites depends not only on the number of satellites and their probability distribution in space , but on the individual values of @xmath91 .
even so , ensemble averaging over many image configurations of the same type ( e.g. quads ) yields an @xmath91 that depends only on the mass of the satellite .
the fractions of quads and doublets that are affected by satellites depend on the fractional area @xmath93 of the lens plane that lies within the circles of influence .
that is , @xmath94 where @xmath95 are different for quads and doublets .
if @xmath93 is small , the circles of influence do not overlap , so @xmath96 the gently increasing @xmath88 in fig .
[ fig : ctm - plummer ] over a range of @xmath82 reflects a dependence @xmath97 for @xmath98 . for @xmath99 , the proportional increase in @xmath83 between , say , @xmath46 and @xmath85 ( see dotted line in fig . [
fig : ctm - plummer ] ) is less than for @xmath81 because of greater overlapping ( @xmath93 increases more slowly than @xmath100 ) .
doublets are much less strongly affected than quads not only because there are only two rather than four images whose fluxes could be affected ( hence the @xmath101 rather than @xmath102 term in ( [ eq:3-f - function - areasat ] ) ) , but , more importantly , because their images are typically of much lower magnification .
let us now allow the total mass @xmath78 in satellite galaxies to vary .
since @xmath88 are not quite independent of how @xmath78 is apportioned between satellites , we need to allow for different individual satellite masses .
we draw them from an @xmath103 distribution with cutoffs at @xmath104 and @xmath105 , and vary @xmath50 to vary @xmath78 .
( cutoff masses of below @xmath106 were too computationally expensive because many more satellites would have been needed for the same total masses . )
the masses of the satellites ( along with their positions ) are regenerated for each new source position .
the results are shown in fig .
[ fig : logplot ] , where @xmath88 are plotted as a function @xmath82 on a log - log scale .
the scalings ( [ eq:3-f - function - areasat ] ) and ( [ eq:3-asat - sum - pireffsqrd ] ) can be seen : @xmath88 initially increase @xmath107linearly with @xmath78 ( as @xmath108 for a given satellite mass function ) , until the overlaps between circles of influence can no longer be neglected , after which @xmath88 approach unity asymptotically .
the proportions @xmath88 go from @xmath109 to @xmath110 over about one order of magnitude of @xmath78 ( this can be seen even more clearly from fig .
[ fig : plummer - sixsets ] in the next section ) .
at all @xmath78 , flux ratios of quads are much more likely to be affected than those of doublets .
= 8.38 cm since the bound mass @xmath111 of a hernquist profile satellite depends , through ( [ eq:2-shq - defn - of - rt ] ) and ( [ eq:2-shq - rs - mt - scaling ] ) on its initial mass @xmath112 and its distance from the main galaxy ( generated randomly in this simulation ) , we can not repeat [ secsub:3-plummer satellites ] exactly .
we can only fix the total initial mass , apportioning it equally between the varying number of satellites @xmath50 .
more - massive satellites , which are therefore less concentrated , are more prone to tidal disruption : for @xmath113 , a total initial mass of @xmath114 results in a total tidally bound mass of @xmath115 if shared equally between 1000 dwarfs of initial mass @xmath46 , and @xmath116 if shared between 10 of initial mass @xmath85 . for @xmath99 these fall to @xmath117 and @xmath118 respectively .
even more significantly , the most massive satellites can not stay bound close - in to the main galaxy ( whereas smaller satellites _ can _ ) , decreasing their chances of lying near the centre of the lens in projection , where the images typically are .
this is dramatically illustrated in fig .
[ fig : ctm - shqd ] , where for @xmath81 the proportion of systems with altered flux ratios only varies weakly with @xmath50 , whereas for @xmath99 satellites with @xmath119 are much more efficient at changing fluxes than massive ones . however , this result depends on the assumption that a satellite which is highly tidally disrupted ( @xmath120 ) can be ignored , on the grounds that its mass surface density is so diffuse that it has little effect on the lensing fluxes .
= 8.38 cm drawing initial masses from the same @xmath103 distribution as before , and varying @xmath50 to vary the total bound mass @xmath78 , we obtain fig .
[ fig : logplot - shqd ] , which is the analogue of fig .
[ fig : logplot ] . the hernquist profile satellites , which are more extended and diffuse than the plummer models , are less efficient at altering fluxes .
= 8.38 cm let us consider several different sets of parameters of elliptical potentials ( [ eq:2-defpiep ] ) and ask what mass in satellites is required for there to be a significant probability of flux - ratio changes .
we first use plummer model satellites , drawn from the @xmath103 mass function , and again vary the number of satellites to vary the total mass .
the results are plotted in figure [ fig : plummer - sixsets ] . for the moderately elliptic case @xmath80 , masses of @xmath121
are needed for 50% of quadruplets to show anomalous flux ratios if @xmath122 decreasing to @xmath123 for @xmath99 .
the lengthscale of the satellite distribution @xmath0 has a large effect on _ both _ @xmath83 and @xmath124 .
as expected , the higher the probability density of satellites in the inner parts of the lens plane ( where the images lie ) , the greater the proportion of anomalous flux ratio systems . comparing left and right panels
, we see that the effect of satellites on flux ratios of quadruplets decreases as @xmath17 increases and the quadruplet cross - section increases , while the mean magnification of quads falls . the proportion of doublets @xmath124 , however , is _ not _ noticeably affected by the ellipticity @xmath125 changing the redshifts of lens and source affects these results only through @xmath126 , on which @xmath127 in eq ( [ eq : wyn ] ) and @xmath128 in eq ( [ eq:2-defplummer - kappa ] ) depend .
@xmath88 for two sets of redshifts are plotted in fig .
[ fig : indep - of - redshift ] , for @xmath99 and @xmath129 .
we see that the results are not very sensitive to different redshifts or different @xmath126 .
= 8.38 cm the results are also not very sensitive to the core radius of the pseudo - isothermal elliptic potential , although they _ are _ affected by its velocity dispersion @xmath15 , as shown in fig .
[ fig : dep - on - veldisp ] .
we recall that the velocity dispersion controls the einstein radius @xmath130 , with more massive lens galaxies diluting the effect of the satellites on flux ratios .
a pseudo - isothermal elliptic potential with @xmath131 requires @xmath132 times as much mass in satellite galaxies as a @xmath133 model for @xmath134 , or 50% of the quads to have anomalous flux ratios .
however , the einstein radius of the main galaxy goes as @xmath135 , and the projected mass within the einstein ring as @xmath136 , so a given _ relative _ mass in satellite companions is more likely to affect flux ratios in more massive elliptical lenses .
= 8.38 cm when the less concentrated , tidally - stripped hernquist profile satellites are used instead of plummer models , the effect on flux ratios is significantly weaker : fig . [
fig : shqd - foursets ] is the analogue of fig .
[ fig : plummer - sixsets ] .
the initial masses of the hernquist dwarfs are drawn from the same @xmath103 distribution , and @xmath88 are plotted against the total tidal mass in satellites . for the @xmath80 elliptical potential , if @xmath81 , some @xmath137 in satellites is needed for @xmath86 of quadruplet flux ratios to be altered ( over three times the mass in plummer satellites needed ) , dropping only to @xmath138 for @xmath99 ( about eight times the mass in plummer satellites needed ) because of the increased tidal stripping of closer - in satellites .
= 8.38 cm = 8.38 cm = 8.38 cm now let us change the lens galaxy to an edge - on spiral , using the model based on eqns ( [ eq : mw - hernquistdensity])-([eq : mw - mn - defpot ] ) .
plots of @xmath88 as functions of @xmath78 , for plummer model satellites , are plotted in fig .
[ fig : mw - versus - piep ] as solid lines alongside those for a pseudo - isothermal elliptic potential as dotted lines .
the upper panel shows the model in which the disc is maximal and provides most of the rotational support in the inner parts , the lower panel shows a model in which the disc is sub - maximal ( see @xcite for a detailed discussion of the lensing properties of these models ) .
there are three solid lines in the panels , as the results are divided according to the five image ( in effect , quadruplet ) , core triplet ( in effect , doublet ) and disc triplet morphologies . compared to a typical elliptical , the flux ratios of maximal disc lens quads are significantly less susceptible to changes through substructure almost three times the mass in satellite galaxies is needed to affect half the flux ratios .
spiral doublets , however , are slightly more susceptible .
this is as expected : the greater the asymmetry of the matter distribution , the larger the 4-image cross - section , the lower the typical magnification of quads and the smaller the effect of satellite galaxies on the 4-image flux ratios .
( so disc triplets , with typical magnifications in between quads and doublets , are also in between in susceptibility to flux changes . ) indeed , we do not expect many anomalous flux ratios in disc galaxy lenses to be caused by the satellite galaxies in their haloes unless these satellites total at least @xmath139 in mass , that is , @xmath140 of the luminous galaxy mass ! ( and _ that _ assumes the satellites are distributed with @xmath99 . ) on moving to the sub - maximal case , the 4-image and disc triplet cross sections are substantially reduced , and so the numbers return towards their values in the elliptical galaxy case .
note that we have compared the effects of a typical spiral with asymptotic circular speed of @xmath141 to a typical elliptical with line - of - sight velocity dispersion of @xmath18 .
this makes sense , as early - type galaxies are more massive than late - type .
if instead we were to carry out the comparison using @xmath142 , so that the kinematic properties of the models were comparable , then the elliptical would have a lower velocity dispersion and so the discrepancy between the effects of spirals and ellipticals would be increased ( see figure [ fig : dep - on - veldisp ] ) . [ cols="^,^,^,^,^,^,^,^ " , ]
it remains unclear whether dark matter satellites and substructure are responsible for anomalous flux ratios in strong lensing .
@xcite originally studied a sample of 7 radio - loud four - image lens systems and claimed evidence of anomalies in 6 of them .
this seemingly suggests that anomalous flux ratios are very common . here ,
we have carried out a theoretical study of the frequency of flux ratio anomalies as a function of lensing galaxy and dark matter substructure parameters .
the likelihood that satellites affect flux ratios in strong lenses depends on their mass profile . here , we considered compact plummer spheres and diffuse , tidally stripped hernquist profiles for our satellites , with the satellite size scaling as a power of mass .
as the hernquist satellites are more extended than the plummer models , they therefore have a smaller effect on image fluxes typically about 3 times the mass is needed to generate the same numbers of anomalous flux ratios .
the probability that strong lensing flux ratios are affected by satellites is crucially dependent on their spatial distribution .
the characteristic lengthscale @xmath0 has a large effect on the fraction of lenses with affected flux ratios .
our spatial distributions of satellite galaxies are inspired by the observational data on the milky way , for which the satellite number density falls off as @xmath143 in three - dimensions with a lengthscale of @xmath144 kpc .
for such distributions , most satellites are too far out to affect the fluxes of images .
for example , even with plummer satellites , a mass of @xmath145 is needed for 20 % of quadruplets to show anomalous flux ratios for a typical elliptical galaxy , rising to @xmath146 for 50 % .
lenses that are edge - on spiral galaxies with maximum discs ( like the milky way ) are more resistant to flux changes by satellites , so the mass in satellites and substructure has to be roughly a factor of 3 times as great for the same proportion of quads to be affected . to obtain anything like the apparent abundance of anomalous flux ratios , then the scalelength of the substructure has to be different to what is known for the milky way satellites .
whether the flux ratios in a lens system are affected by satellites is sensitive to the total mass in satellites , but more weakly dependent of how this mass is apportioned between them . for plummer model satellites , the probability that a given satellite changes a flux ratio increases with its mass only slightly faster than linearly , at least when its mass is between @xmath147 and @xmath148 .
for example , satellites of mass @xmath149 are only responsible for the causing @xmath150 more flux ratio anomalies than those of mass @xmath151
. for hernquist model satellites , more massive ones seem no more efficient ( per unit mass ) at changing fluxes ; indeed , more massive dwarfs , being more prone to tidal disruption , might even be _ less _ efficient than lighter , more compact ones .
one interesting consequence is that , if matter in dark satellites is not predominantly in the most massive ones , then the contribution of the most massive satellite galaxies to flux ratio anomalies should not be predominant . in the light of this
, the fact that so many anomalous flux ratios systems have optically identified substructure seems at outset surprising .
there seem to be two possible explanations .
first , the visible substructure may have been wrongly identified as the cause , whereas dark substructure may be the true culprit .
a large dwarf galaxy may by chance be projected close to the einstein radius , whereas unrelated dark substructure may be the major cause of the anomaly .
second , visible satellites may be more concentrated than their dark cousins , a physical effect that may naturally arise from baryon condensation .
compression factors causing an enhancement of the central density by a factor of 10 in bright satellites seem to be ample to give a satisfactory explanation of the observed statistics . nonetheless , such high compression factors are probably implausible except for the largest satellite galaxies .
this seems to be in accord with the results of @xcite , who found that including baryons in numerical simulations did not help in reconciling simulation results with the statistics of anomalous flux ratios . finally , we remark that the likelihood that flux ratios are affected depends on the ellipticity of the main lens galaxy , but this dependence is much stronger in quads than doublets .
this is a consequence of high magnification images being more easily affected by a dwarf galaxy than low magnification ones .
quads are more highly magnified than doublets ( the disc triplets of spirals are in between ) , and changing the ellipticity of the main galaxy changes the typical magnification of quads more than it changes that of doublets .
generally , the greater the ellipticity the less the effect of satellite galaxies . for example , a given mass @xmath78 of dwarf satellites around a typical edge - on maximum - disc spiral galaxy is significantly less likely to change image flux ratios than @xmath78 around a typical elliptical galaxy , even though the spiral is less massive than the elliptical .
we thank the referee for some helpful comments .
ems thanks the commonwealth scholarship commission and the cambridge commonwealth trust for the award of a studentship .
nwe thanks neal jackson for some insightful discussions .
this work has been supported by the angles network . | gravitational lenses with anomalous flux ratios are often cited as possible evidence for dark matter satellites predicted by simulations of hierarchical merging in cold dark matter cosmogonies .
we show that the fraction of quads with anomalous flux ratios depends primarily on the total mass and spatial extent of the satellites , and the characteristic lengthscale @xmath0 of their distribution . if @xmath1 , then for a moderately elliptical galaxy with a line - of - sight velocity dispersion of @xmath2 , a mass of @xmath3 in highly - concentrated ( plummer model ) satellites is needed for 20% of quadruplets to show anomalous flux ratios , rising to @xmath4 for 50% .
several times these masses are required if the satellites have more extended hernquist profiles .
compared to a typical elliptical , the flux ratios of quads formed by typical edge - on disc galaxies with maximum discs are significantly less susceptible to changes through substructure three times the mass in satellite galaxies is needed to affect 50% of the systems . in many of the lens systems with anomalous flux ratios , there is evidence for visible satellites ( e.g. , b2045 + 265 or mg0414 + 0534 ) .
we show that if the anomaly is produced by substructure with properties similar to the simulations , then optically identified substructure should not be preponderant among lens systems with anomalies .
there seem to be two possible resolutions of this difficulty .
first , in some cases , visible substructure may be projected within or close to the einstein radius and wrongly ascribed as the culprit , whereas dark matter substructure is causing the flux anomaly .
second , bright satellites , in which baryon cooling and condensation has taken place , may have higher central densities than dark satellites , rendering them more efficient at causing flux anomalies .
[ firstpage ] gravitational lensing
dark matter |
nearby galaxy clusters exhibit a tight correlation between x - ray isophotal size and emission weighted intracluster medium ( icm ) temperature ( @xcite ; hereafter me97 ) .
this correlation is evidence of regularity ; it exists in an x - ray flux limited sample of 45 clusters ( @xcite ) , where no attempt has been made to use the x - ray morphologies to exclude clusters showing signatures of recent , major mergers .
the scatter around the x - ray size temperature ( st ) relation is approximately 15% in size , comparable to the scatter of elliptical and lenticular galaxies around their fundamental plane ( @xcite ) .
this small scatter in the galaxy cluster scaling relation is intriguing , because ( 1 ) there is overwhelming evidence that galaxy clusters are still accreting mass ( e.g. @xcite ) and ( 2 ) elliptical galaxies are generally thought to be among the most regular objects in the universe .
me97 use 48 n - body and hydrodynamical simulations of cluster formation in four different cosmological models to address this apparent contradiction between regularity and ongoing accretion in nearby clusters . using simulations from both @xmath8 and @xmath4 cosmologies ,
they show that a tight st relation is expected even in cosmogonies where there is significant cluster growth at the present epoch .
the high degree of regularity implied by the st relation is surprising , because the well known correlation between x - ray luminosity and emission weighted mean temperature ( the relation ) has very large scatter ( @xcite ) .
me97 show that the same cluster ensemble which exhibits a 15% scatter in the st relation exhibits a 52% scatter in @xmath9 around the relation .
this higher scatter in the relation results from the sensitivity of the x - ray luminosity to the densest regions of the cluster a sensitivity to the presence or absence of so - called cooling flows ( @xcite ) .
this interpretation is supported by more recent work where cluster ensembles specially chosen to contain no cooling flow clusters conform to relations with significantly reduced scatter of 25% ( @xcite ) . additionally , when central parts of cooling flow clusters are excluded , the scatter in the relation decreases ( @xcite ) .
finally , cluster regularity is also evident in the tight correlation between icm mass and emission weighted temperature ( the relation ) . when measuring icm mass within a limiting radius of ( the radius where the enclosed overdensity is 500 times the critical density ) the scatter in mass about the relation is 17% ( @xcite ) .
observational studies of the st relation followed by work on the and relations support a scenario where clusters exhibit regularity similar to that of elliptical galaxies in the properties measured on the scales of their virial regions , but exhibit significant irregularities in the properties of the densest , central regions where physical processes other than gravity and gas dynamics such as radiative cooling and magnetic fields play significant roles ( @xcite ) .
the evidence indicating cluster regularity is balanced by evidence for departures from regularity ; the scatter in the observed scaling relations is larger than can be accounted for by the measurement uncertainties .
this resolved scatter contains clues about , among other things , cluster peculiar velocities and departures from equilibrium .
here we examine the st relation at intermediate redshift ( @xmath10 ) using rosat hri observations of the canadian network for observational cosmology ( cnoc ; e.g. @xcite ) cluster sample .
the st relation provides a potentially powerful tool to study the expansion history of the universe . as explained in detail below , the st relation is rather insensitive to cosmological evolution .
thus , armed with an accurate model of cluster evolution , one could use the st relation to measure distances at intermediate redshift , constraining the deceleration parameter @xmath11 .
we first present the x - ray st relation in the nearby cluster sample ( @xmath12[sec : localst ] ) and then present an interpretation of the st relation in terms of regularity in the underlying dark matter properties of the cluster .
section [ sec : midzst ] describes observations of the st relation in intermediate redshift clusters . in @xmath12[sec : cosmo ] we use these observations to constrain cosmological parameters .
section [ sec : conclude ] contains a summary of our conclusions . throughout the paper
we use @xmath13 .
below we present observations of the st relation in an x - ray flux limited sample of clusters and then discuss cluster scaling relations .
fig [ fig : stnearby ] contains a plot of the x - ray st relation for an ensemble of nearby galaxy clusters ( see me97 for cluster list ) .
this sample is essentially a low redshift sample , @xmath14 with rms variation @xmath15 , but the full redshift range is @xmath16 .
the cluster isophotal size @xmath17 is plotted versus the emission weighted mean icm temperature , where @xmath18 and @xmath19 is the area enclosed by the isophote @xmath20 . the isophotal size is a model - independent measure extracted directly from the x - ray image .
icm temperatures come from the literature ( listed in me97 ) . in this figure
we use the galactic absorption corrected surface brightness @xmath21 within the cluster rest frame 0.5 - 2.0 kev band . in the conversion from count rate to physical flux units we use pros , assume a raymond - smith spectrum with the published mean temperature @xmath22 and @xmath23 cosmic abundances ( @xcite ) .
we convert between the angular size and physical size of the isophotal region using the angular diameter distance , and we also correct @xmath20 for cosmological dimming @xmath24 .
the best fit power law to this local st relation has the form @xmath25 and the scatter about this relation is 15% in @xmath17 .
this fit is shown as a solid line in fig [ fig : stnearby ] . because this is the local st relation , it is only weakly dependent on the deceleration parameter . for this relation
we have assumed the parameters @xmath8 and @xmath26 ; for @xmath8 and @xmath5 the normalization is 1.8% lower , and for @xmath4 the normalization is 2.7% lower .
galaxy cluster scaling relations between , for example , virial mass and temperature @xmath27 are expected if clusters are approximately virialized .
the spherical collapse model predicts that newly collapsed objects will have mean densities which are @xmath28 times the critical density @xmath29 , where @xmath30 is the hubble parameter and in an einstein - de sitter model @xmath31 .
therefore , we define the virial radius @xmath32 of an object with virial mass @xmath33 to be @xmath34 newly collapsed clusters may also satisfy the virial relation @xmath35 where @xmath27 is the virial temperature and @xmath36 is a number dependent on the cluster density and temperature structure ; in the case that cluster structure is self similar ( i.e. that massive clusters are simply rescaled versions of low mass clusters ) @xmath36 will be constant for all @xmath27 and eqns [ eq : rdelta ] and [ eq : virial ] lead to the well known scaling relations between virial mass , radius and temperature @xmath37 where @xmath38 is a dimensionless structure parameter like @xmath36 .
these scaling relations are expected to hold for a range of @xmath39 and @xmath40 , where @xmath39 is the region with enclosed overdensity @xmath41 with respect to @xmath42 and @xmath40 is the corresponding enclosed mass .
numerical cluster simulations indicate that the emission weighted mean icm temperature is an adequate proxy for the virial temperature @xmath27 , and that these scaling relations exist even in cluster populations which exhibit merging and substructure ( @xcite ) . using simulations to normalize these relations
, one finds that clusters with of 10 kev have virial masses @xmath43 and virial radii @xmath44 .
the small scatter in the observed st relation , the correlation between icm mass and temperature ( see section 1 ) , and a direct study of the correlation between binding mass ( estimated using cluster galaxy kinematics ) and emission weighted temperature ( @xcite ) provide evidence that the conditions required for eqn [ eq : scaling ] to be valid are generally met in real galaxy clusters .
the x - ray surface brightness ( units : ergs s@xmath45@xmath46sr@xmath45 ) at a projected radius @xmath47 from the cluster center is @xmath48 where @xmath49 and @xmath50 are the electron and proton number densities , @xmath51 is the distance from the cluster center and @xmath52 is the x - ray emission coefficient which includes contributions from thermal bremsstrahlung and line emission ( @xcite ) .
we can express the icm density @xmath53 in terms of the underlying dark matter density @xmath54 : @xmath55 , where @xmath56 is the icm mass fraction within the virial region and @xmath57 is a function describing differences in the dark matter and icm density distributions ; then , using @xmath58 , we can rewrite the x - ray surface brightness as @xmath59 we bring the emission coefficient out of the integral and use the value at @xmath60 ; although this is only strictly valid for an isothermal gas , the temperature insensitivity of @xmath52 band limited to 0.5:2.0 kev makes this an excellent approximation even in the presence of departures from isothermality ( e.g. @xcite ) .
finally , we express the dark matter density profile in terms of the characteristic overdensity of the virial region : @xmath61 , where @xmath62 encodes the dependence of the dark matter profile on a dimensionless radius @xmath63 ; this approach is consistent with numerical cluster simulations , which indicate that cluster dark matter density profiles have a `` universal '' form ( @xcite , hereafter nfw ) .
more recently , higher resolution simulations have shown differences between the inner profile ( at radii which are 1% of the virial radius ) and the form proposed by nfw ( @xcite ) ; these differences appear to be smallest for cluster scale haloes , and none of our conclusions are sensitive to the behavior of the density profile at 1% of the virial radius . expressing @xmath54 in this way , and removing the characteristic scale @xmath47 from the integral , the x - ray surface brightness becomes @xmath64 where @xmath65 is a dimensionless integral which encodes the shape of the icm density profile : @xmath66 , where @xmath67 and @xmath68 .
we relate @xmath47 and @xmath17 using the shape of the typical x - ray surface brightness profile ; @xmath69 is typically well fit by the so - called @xmath70 model ( @xcite ) @xmath71 where @xmath72 is the core radius ; well outside the core the surface brightness falls as @xmath73 .
therefore we write @xmath74 using eqn .
[ eq : scaling ] to substitute for @xmath47 , one can readily determine the st relation at a particular redshift in the case that cluster structure is self similar @xmath75 for @xmath76 , a typical observed value ( @xcite ) , the predicted st relation has @xmath77 , shallower than the observed relation : @xmath78 .
the observed increase in icm mass fraction @xmath56 with cluster temperature ( @xcite ) is enough to explain the steepness of the observed relation .
the physics underlying this systematic variation in cluster gas mass fraction is likely heating of the intergalactic medium during the process of galaxy formation .
this so - called preheating affects the structure of the gas in low mass clusters more so than in high mass clusters ( e.g. @xcite ) .
as discussed below , in comparing the local and intermediate redshift st relations it is important only that the effects of this preheating be similar in both samples .
before presenting the intermediate redshift st relation , we present an evolution model and our measurement methods .
comparison of st relations in nearby and distant populations requires a model for how the expected changes in cluster structure with increasing redshift will affect the st relation . in the previous section we presented an explanation of the correlation between cluster
x ray isophotal size and emission weighted icm temperature as a manifestation of underlying scaling relations in the dark matter properties of the cluster .
following this line of reasoning , we now express the evolution of the st relation in terms of the redshift evolution of these same dark matter properties .
the evolution of the dark matter scaling relations is apparent in eqn [ eq : scaling ] .
the normalization of these scaling relations changes with redshift according to the change in @xmath42 and the change in the characteristic overdensity @xmath28 of collapsed haloes . within einstein - de
sitter models @xmath28 does not evolve , whereas evolution is expected in low @xmath2 models .
for simplicity of presentation , in the following analysis we explicitly follow only changes in @xmath42 , but accounting for changes in @xmath28 would not affect our conclusions . using @xmath79 , we express the evolution of the mass and size scaling relations as @xmath80 where @xmath81 , and @xmath2 and @xmath3 are the present epoch contributions to the density parameter from matter and `` dark energy '' , respectively .
these simple evolution models are valid only for self similar evolution i.e. in the case that distant cluster dark matter profiles are structurally similar to those of nearby clusters . from eqn [ eq : idelta ]
it follows that the x - ray surface brightness at @xmath47 evolves as @xmath82 and that the normalization of the st relation evolves as @xmath83 interestingly , for the most common cluster profile @xmath76 ( local : @xcite ; this intermediate redshift sample : @xmath84 ) , so @xmath85 , corresponding to no evolution in the st relation .
in essence , a cluster of a given is denser in the past , boosting its x - ray surface brightness and tending to increase the isophotal radius @xmath17 , but this cluster is also smaller in the past , tending to decrease the isophotal radius @xmath17 .
if the icm density profile behaves as @xmath86 , these two effects cancel .
thus , a population of clusters with measured @xmath87 will exhibit an st relation which is relatively insensitive to cosmological evolution . in principle
, this behavior makes the st relation ideal for measuring angular diameter distances to high redshift clusters .
we use archival rosat hri observations of 11 of the 14 cnoc clusters discussed in lewis et al .
( 1999 ) ; these 11 clusters are those with published measurements .
a detailed description of the reduction to 0.5@xmath88 pixel images is given in lewis et al .
the hri angular resolution is @xmath89 fwhm ; we further bin these images to a final pixel size of 2@xmath88 without significant loss of angular resolution . despite the far higher median exposure time
( 33 ks compared to 8 ks ) , the image quality of these intermediate redshift clusters is far lower than for our nearby sample ( due to combination of lower observer frame surface brightness and higher instrument background ) .
the image quality is too poor to allow isophotal sizes measurements directly from the images using the approach described in @xmath12[sec : localst ] ; therefore , we fit circular @xmath70 models ( same form as eqn [ eq : beta ] with @xmath90 substituted for @xmath91 ) to these images using the software developed for measuring sze+x - ray distances ( @xcite ) .
essentially , we find the set of parameters @xmath92 , @xmath93 , @xmath70 , cluster centroid @xmath94 and local background @xmath95 which maximizes the likelihood of consistency between model and data .
in all these fits we fix the background to the value measured in an annulus extending from 4.5@xmath96 to 5@xmath96 , a background dominated region with an essentially negligible contribution from the cluster .
fits are performed to the central portion ( @xmath97 ) of the image which includes the cluster and a local background region . in table
[ tab : fitparam ] we list the best fit parameters with estimates of the statistical uncertainties .
the central surface brightness and measured background are both given in detector units of cts s@xmath45arcmin@xmath98 .
these cluster parameters are in reasonably good agreement with those presented in lewis et al .
( 1999 ) , with the exception of two clusters a2390 and ms1358 .
the different best fit parameters in these two clusters stem from the treatment of emission excesses ; we fit the surface brightness model to the entire dataset , whereas lewis et al .
( 1999 ) exclude the central region in both these clusters .
this exclusion approach typically leads to larger core radii and correspondingly higher @xmath70 s than would fitting to the entire cluster as in our method .
an examination of the fit residuals in our sample provides no clear evidence for systematic differences between the data and the models with the exception of a2390 , which exhibits an asymmetry or centroid variation ( e.g. @xcite ) .
allowing for an ellipticity introduces two additional free parameters , and the information content of some of the images simply is not sufficient to provide meaningful constraints .
our goal here is to determine whether there is any evidence of regularity in intermediate redshift clusters ; any mismatch between the data and the model ( like that in a2390 ) will serve simply as an additional source of scatter in the final relation . therefore , we conservatively present a uniform and objective analysis which is appropriate for the majority of our sample , rather than varying our analysis from cluster to cluster .
( an elliptical @xmath70 model fit to a2390 yields an isophotal size which is 8% smaller than that listed in table [ tab : fitparam ] .
this correction makes a2390 more consistent with the best fit st relation , but the correction is unimportant compared to the 17% rms scatter of the sample about that best fit relation . ) of course the best approach is a non - parametric analysis of the surface brightness maps such as that applied to the local sample , but this is not possible with the current data .
we use pros to convert the detector count rate into galactic absorption corrected flux units of ergs s@xmath45@xmath46 within the cluster rest frame 0.5:2.0 kev band . as for the low redshift clusters , we assume a raymond - smith emission spectrum with the measured mean temperature @xmath22 and @xmath23 cosmic abundances ( @xcite ) at the appropriate redshift , and then we correct for the @xmath24 cosmological dimming . the rest frame 0.5:2.0 kev central surface brightnesses in units of ergs s@xmath45@xmath46arcmin@xmath45
are also listed for each cluster .
lccccccccr a2390 & 0.2279 & 8.9@xmath99 & 0.159@xmath100 & 9.33 & 17.2@xmath101 & 0.532@xmath102 & 0.0022 & 235.1@xmath103 & 1.202 ms0015.9 + 1609 & 0.5466 & 8.0@xmath104 & 0.018@xmath105 & 1.63 & 49.4@xmath106 & 0.898@xmath107 & 0.0020 & 112.3@xmath108 & 1.005 ms0440.5 + 0204 & 0.1965 & 5.3@xmath109 & 0.079@xmath110 & 4.71 & 9.1@xmath111 & 0.521@xmath112 & 0.0023 & 97.7@xmath113 & 0.445 ms0451.5 + 0250 & 0.2010 & 8.6@xmath104 & 0.012@xmath105 & 0.69 & 94.6@xmath114 & 0.750 & 0.0024 & 211.0@xmath115 & 0.979 ms0451.6 - 0305 & 0.5392 & 10.4@xmath116 & 0.025@xmath117 & 2.30 & 36.4@xmath118 & 0.752@xmath119 & 0.0023 & 119.7@xmath120 & 1.064 ms0839.8 + 2938 & 0.1928 & 4.2@xmath121 & 0.129@xmath122 & 6.33 & 11.9@xmath123 & 0.588@xmath124 & 0.0023 & 98.4@xmath125 & 0.442 ms1008.1 - 1224 & 0.3062 & 7.3@xmath126 & 0.018@xmath105 & 1.24 & 35.8@xmath127 & 0.667@xmath128 & 0.0023 & 118.5@xmath129 & 0.745 ms1224.7 + 2007 & 0.3255 & 4.3@xmath130 & 0.054@xmath131 & 3.19 & 7.5@xmath132 & 0.552@xmath133 & 0.0025 & 55.9@xmath134 & 0.368 ms1358.4 + 6245 & 0.3290 & 6.5@xmath135 & 0.111@xmath136 & 6.14 & 8.4@xmath137 & 0.501@xmath138 & 0.0023 & 119.7@xmath139 & 0.795 ms1455.0 + 2232 & 0.2570 & 5.5@xmath140 & 0.460@xmath141 & 31.86 & 11.2@xmath142 & 0.643@xmath143 & 0.0028 & 127.7@xmath144 & 0.713 ms1512.4 + 3647 & 0.3726 & 3.6@xmath145 & 0.107@xmath146 & 6.20 & 7.4@xmath147 & 0.560@xmath148 & 0.0022 & 70.1@xmath149 & 0.505 [ tab : fitparam ] angular isophotal sizes @xmath150 are then estimated using a fiducial surface brightness of @xmath151 ergs s@xmath45@xmath46arcmin@xmath98 .
we determine statistical uncertainties in the fit parameters by exploring the likelihood within a grid in the parameters @xmath92 , @xmath93 and @xmath70 ; we then use that range to estimate uncertainties in the derived isophotal size @xmath150 .
figure [ fig : uncertain ] contains a plot of the 1 , 2 and 3@xmath6 confidence regions in @xmath70 and @xmath93 for ms1455 ; lines of constant isophotal size @xmath150 are overlaid .
note that lines of constant @xmath150 are approximately parallel to the well known @xmath152 correlation ( e.g. @xcite ) , minimizing the detrimental effects this correlation has on our size measurement uncertainties .
measured @xmath150 are then converted into physical isophotal sizes @xmath17 using the angular diameter distance @xmath153 : @xmath154 .
we use the general form @xmath155 where @xmath156 , @xmath157 , and @xmath158 in an open , flat and closed cosmology .
table [ tab : fitparam ] contains a list of @xmath150 for @xmath159 ergs / s/@xmath160/arcmin@xmath161 with statistical uncertainty estimates and the corresponding @xmath17 for the case where @xmath8 & @xmath26 .
figure [ fig : stmidz ] contains the intermediate redshift st relation with 1@xmath6 error bars in both @xmath22 and @xmath17 .
the conversion from @xmath150 to @xmath17 assumes @xmath8 and @xmath26 in this figure .
the best fit relation in this case is @xmath162 and the rms scatter in size about this relation is 17% .
also plotted is the local st relation for this same isophote ( solid line ) .
the uncertainties are derived by bootstrap resampling the list of 11 sizes ( allowing duplication ) and refitting the relation .
there is a suggestion that the intermediate redshift st relation is steeper than the local relation , but the difference in slopes is less than 1@xmath6 significant .
in fact , both the zeropoints and slopes of the local and intermediate st relations are statistically consistent when we use this particular set of cosmological parameters . with a larger sample
it will be possible to measure the intermediate redshift st relation slope more accurately ; a comparison of the two slopes would be an important diagnostic of unusual structural evolution in clusters .
figure [ fig : stmidzscdm ] contains a plot of the intermediate redshift st relation assuming @xmath4 .
the best fit slope in this case is @xmath163 , and the zeropoint is @xmath164 .
this zeropoint differs from that of the local st relation ( also calculated with @xmath4 ) at 2.5@xmath6 .
we have examined the local and intermediate redshift st relations at four other isophotes : @xmath165 cgs , @xmath166 cgs , @xmath167 cgs , and @xmath168 cgs .
the general conclusions reached above using the isophote @xmath151 cgs apply equally as well at these other isophotes : there is a suggestion that the st relation at intermediate redshift is steeper ( at less than 1 @xmath6 significance ) than the local relation , and the zeropoints are in good agreement when using the cosmological parameters @xmath8 and @xmath26 .
the fainter the isophote the larger the enclosed region . pushing to fainter isophotes is dangerous given the quality of the intermediate redshift data , and pushing to brighter isophotes leads to complications from central emission excesses and the effects of the psf .
thus , we choose the isophote @xmath151 cgs as a compromise between these two competing effects .
although the x - ray images of these 11 clusters are poor , it is nevertheless interesting to use the local st relation and our evolution model to predict distances to the intermediate redshift sample , thereby constraining cosmological parameters . because the clusters are at a range of redshifts , both the scatter of the data about the local st relation and any systematic offsets from the local relation contain cosmological information .
operationally , for each set of cosmological parameters @xmath2 and @xmath3 we convert the measured @xmath169 and calculate the @xmath170 of the sample about the best fit local st relation , where the best fit local relation is calculated using this same set of cosmological parameters .
fig [ fig : cosmo ] contains contours of @xmath171 corresponding to 1 , 2 and 3 @xmath6 confidence regions ( equivalent to @xmath172 , 6.2 and 11.8 ) in the @xmath2@xmath3 space . the current sample ,
because of observational uncertainties , does not provide a strong cosmological constraint .
the 1@xmath6 confidence region is a wide trough similar in character to those derived from luminosity distances to sne ia ( @xcite ) , with the important difference that our sample of 11 intermediate redshift distances is not as constraining as the larger samples of sne ia distances .
notably , there is enough power in the current dataset to exclude the @xmath4 & @xmath5 model ( one of three models marked with stars ) at @xmath17395% confidence . if one considers only models where @xmath7 , then the preferred model has @xmath174 .
furthermore , @xmath175 with 1@xmath6 confidence and @xmath176 with 3@xmath6 confidence .
one can also use these data to constrain the deceleration parameter @xmath177 .
the preferred value is @xmath178 ( 68% confidence ) .
clearly , larger samples of archival data and the impending stream of high quality x - ray images from chandra and xmm will provide tighter constraints .
one particular concern in this study is that the local and intermediate redshift samples were observed with two different instruments : the rosat pspc and the rosat hri .
any relative calibration errors would then serve to bias the preferred cosmological parameters .
however , the steepness of cluster x - ray surface brightness profiles mitigates this potential problem ; typically , the surface brightness falls off as @xmath179 , so even a 10% relative error in the calibration of the pspc and hri would introduce only a 3% error in the angular size of the cluster .
although there has been no previous use of cluster scaling relations to constrain cosmological parameters , there have been three attempts to use the icm mass fraction ( @xcite ) ; if one assumes is constant with redshift , then one can use its distance dependence to constrain the distance redshift relation .
the pen constraint @xmath180 is marginally inconsistent with our measurement , but the pen analysis is only appropriate in the case that the icm traces the dark matter , a case that is inconsistent with both theoretical and observational studies ( @xcite ) . in the rines
analysis , care is taken to ( 1 ) match the luminosity ranges of the local and intermediate redshift samples , ( 2 ) address the expected radial variation of , and ( 3 ) test the consistency of the published effective areas of the two satellites involved ( einstein and asca ) . because increases slowly with radius ,
it is important that it be measured within the same portion of the virial region in the local and intermediate redshift samples ; this complicates the use of measurements as cosmological constraints . if one analyzes local and intermediate redshift clusters at a fixed metric aperture , the region studied will correspond to a larger fraction of the virial region as the redshift increases ( see eqn .
[ eq : scaleevolution ] ) .
similarly , when measuring within a fixed portion of the virial region , it is important that the redshift evolution of the @xmath181 be included ; it appears that in both the rines et al . and cooray analyses , no evolution was included , so the portion of the virial region used in determining increases with redshift . in both cases ,
these biases cause the inferred angular diameter distances to be underestimated .
the size of the bias depends on the cosmological model being evaluated ( eqn .
[ eq : scaleevolution ] ) ; using the relation @xmath182 ( @xcite ) , we estimate the bias on @xmath153 at @xmath183 to be -2% for @xmath8 & @xmath26 , and -10% for @xmath4 and @xmath5 ; these biases are important relative to the @xmath1848% difference in @xmath185 between an @xmath8 and an @xmath8 & @xmath26 cosmology . correcting for this bias will improve the consistency between our cosmological constraints and those of rines et al . and cooray .
we present the local x - ray size temperature ( st ) relation for an x - ray flux limited sample of 45 clusters observed with the rosat pspc ( @xcite ) .
we provide an explanation of this scaling relation in terms of underlying scaling relations in the cluster dark matter properties .
the observed st relation is slightly steeper than the self
similar expectation presented in @xmath12[sec : stscaling ] , but is consistent when the modest variation of the icm mass fraction @xmath56 with @xmath186 is taken into account ( @xcite ) .
we use our theoretical model for the st relation to explore its evolution with redshift .
interestingly , for the typical icm radial distribution observed in nearby and distant clusters , the normalization of the st relation is not expected to evolve .
the lack of evolution makes the st relation a plausible source of intermediate redshift angular diameter distances . of course ,
if cluster structure evolution is very different from the current theoretical expectation , our model ( @xmath12[sec : stevolve ] ) will underestimate st relation evolution .
for example , a shift in the mean icm mass fraction with redshift would bias st relation distances ; to date there is no compelling evidence that distant clusters have different icm mass fractions in the mean than nearby clusters ( e.g. @xcite ) .
we use rosat hri observations of 11 cnoc clusters with measured emission weighted mean icm temperatures to make the first measurements of the intermediate redshift st relation . because of the poor image quality
, we measure the angular isophotal size @xmath150 using the best fit circular @xmath70 model , rather than measuring it nonparametrically as for the nearby clusters ( see eqn [ eq : size ] ) . by assuming the cosmological parameters @xmath8 & @xmath26 , we show that the slope and zeropoint of this intermediate redshift st relation is statistically consistent with that of the local st relation ( see fig [ fig : stmidz ] ) .
in addition , we examine the relation for @xmath4 & @xmath5 ( see fig [ fig : stmidzscdm ] ) , showing that although the slope is consistent , the zeropoint is different at greater than 3@xmath6 significance .
finally , we use this cluster sample and our st relation evolution model to place cosmological constraints . given the quality of the cluster images and temperature measurements ,
it is not surprising that a wide range of cosmological models is consistent with the data .
nevertheless , this sample of 11 intermediate redshift distances is sufficient to rule out @xmath4 with between 2@xmath6 and 3@xmath6 confidence . taken together with icm mass fraction constraints on the cosmological matter density parameter @xmath187 at 95% confidence ( @xcite )
, the cluster st relation exhibits a slight preference for universes with @xmath188 ; models with @xmath5 are inconsistent with the st relation at between 1 and 2@xmath6 . when considering only models where @xmath7 , we can rule out @xmath4 with 3@xmath6 confidence
. with the higher quality x - ray images and icm temperature measurements available from chandra and xmm , a significant tightening of these constraints and further tests of the underlying evolution model will be possible .
comparison of local and distant relations , which are more sensitive to cluster evolution , would provide important constraints on these models .
in addition , observations with a new generation of sunyaev - zeldovich effect instruments ( @xcite ) will allow us to more accurately constrain the evolution of cluster structure . with these future observations of intermediate and high redshift clusters ,
we plan to continue using the st relation as a tool to provide cosmological constraints independent of those derived from recent high redshift sne ia observations ( @xcite ) .
jjm is supported by chandra fellowship grant pf8 - 1003 , awarded through the chandra science center .
the chandra science center is operated by the smithsonian astrophysical observatory for nasa under contract nas8 - 39073 .
edr is supported by nasa gsrp fellowship ngt5 - 50173 .
ee acknowledges support provided by the national science foundation grant ast 9617145 .
aee acknowledges support from nsf ast-9803199 and nasa nag5 - 8458 .
this research has made use of data obtained through the high energy astrophysics science archive research center online service , provided by the nasa / goddard space flight center . | we present the first measurements of the x - ray size - temperature ( st ) relation in intermediate redshift ( @xmath0 ) galaxy clusters .
we interpret the local st relation ( @xmath1 ) in terms of underlying scaling relations in the cluster dark matter properties , and then we use standard models for the redshift evolution of those dark matter properties to show that the st relation does not evolve with redshift .
we then use rosat hri observations of 11 clusters to examine the intermediate redshift st relation ; for currently favored cosmological parameters , the intermediate redshift st relation is consistent with that of local clusters .
finally , we use the st relation and our evolution model to measure angular diameter distances ; with these 11 distances we evaluate constraints on @xmath2 and @xmath3 which are consistent with those derived from studies of type ia supernovae . the data rule out a model with @xmath4 and @xmath5 with 2.5@xmath6 confidence . when limited to models where @xmath7 , these data are inconsistent with @xmath4 with 3@xmath6 confidence . |
the layered ruthenium oxides sr@xmath1ruo@xmath4 , sr@xmath0ru@xmath1o@xmath2 , and the perovskites srruo@xmath0 and caruo@xmath0 , have attracted much attentions during the last several years primarily because these systems allow quantum phase transitions in clean metals @xcite to be addressed experimentally .
the appearances of novel transitions associated with the correlated 4@xmath3 electrons continues to stir up significant issues within the condensed matter community .
the layered member sr@xmath1ruo@xmath4 , for example , enters a superconducting state at about 1 k in association with a rather unusual spin - triplet type electron - pairing , clearly at odds with predictions of conventional @xmath5wave superconductivity @xcite .
the bilayered member , sr@xmath0ru@xmath1o@xmath2 , shows non - fermi liquid transport in the vicinity of a metamagnetic transition at high magnetic field and low temperature , suggesting that the scattering rate of the conduction electrons may be influenced by quantum fluctuations , as opposed to more conventional thermal fluctuations @xcite .
finally , data suggest that a quantum critical point exists in the solid solution sr@xmath6ca@xmath7ruo@xmath0 for @xmath80.7 between ferromagnetic ( sr - side ) and nearly ferromagnetic ( ca - side ) ordered states .
these materials have provided experimental opportunities to study criticality of correlated electrons in the vicinity of the quantum critical points @xcite .
the present status of investigations of quantum phase transitions and criticality seem far behind that of the conventional transitions that are driven by thermal fluctuations @xcite . in order to promote progress in the studies of quantum phase transitions and criticality in clean metals , and to answer many open questions aroused thus far , we need further opportunities to study these issues experimentally . searching for a new class of materials which show quantum critical behavior within realizable magnetic , pressure , or chemical ranges , should be encouraged .
high - quality single crystals of these materials , if available , would be highly desirable for experimental investigations .
we have been synthesizing the variety of compounds in the sr - rh(iv ) oxide system , where rh is expected to be 4@xmath9 , by applying high - temperature and high - pressure techniques at typically 6 gpa and 1500 @xmath10c @xcite . besides the counted members of the strontium - rhodium system ,
the perovskite - type compound was recently obtained by quenching to room temperature at elevated pressure @xcite .
the formation of srrho@xmath0 was an excellent example of what can be achieved by high - pressure techniques no compounds at the ratio sr : rh = 1:1 had ever been reported until the high - pressure experiment . while the sr substitution for larh@xmath11o@xmath0 was studied by a regular solid - state - reaction technique at ambient pressure ; it was then found that the solubility limit of sr at the la site was at most 10 % , and the solid solution remained semiconducting in the very narrow region @xcite .
it is then clear that the formation of the 100 % end - member srrho@xmath0 goes far beyond the limit achieved by the regular synthesis study , and it is indicative of the dramatic effectiveness of the high - pressure experiment .
this synthesis technique can be expected to uncover further novel electronic materials in addition to srrho@xmath0 .
the magnetic and electrical properties of srrho@xmath0 were qualitatively similar of those observed in the analogous ruthenium oxide caruo@xmath0 , the latter being a nearly ordered ferromagnetic metal that is probably influenced by spin - fluctuations @xcite .
we had therefore expected unusual electronic properties in the unforeseen compounds in the rhodium oxide system , as well as the analogous ruthenium oxides .
we have focused our attention to obtaining further novel materials in the sr - rh - o space , and thus far , the ruddlesden - popper - type member sr@xmath0rh@xmath1o@xmath2 was found . in this paper
we present data from neutron diffraction study , magnetic and electrical measurements on a polycrystalline sample of this new phase , and the results are compared to the ruthenium analogue .
the polycrystalline sample of sr@xmath0rh@xmath1o@xmath2 was prepared as follows .
the fine and pure ( @xmath12 ) powders of sro@xmath1 , rh@xmath1o@xmath0 , and rh were mixed at the compositions sr@xmath0rh@xmath1o@xmath13 , where @xmath14 0.0 and 0.1 .
approximately 0.2 grams of each stoichiometry was placed into a platinum capsule and then compressed at 6 gpa in a high - pressure apparatus , which was originally developed in our laboratory @xcite .
the sample was then heated at 1500 @xmath10c for 1 hr and quenched to room temperature before releasing the pressure .
the quality of the products was investigated by means of powder x - ray diffraction ( cuk@xmath15 ) .
the x - ray profile was carefully obtained at room temperature using the high - resolution - powder diffractometer ( rint-2000 system , developed by rigaku , co ) , which was equipped with a graphite monochromator on the counter side .
we found that the both samples were of high quality ; there were no significant impurities in either product , and no remarkable difference between them .
the x - ray pattern for the powder sample at @xmath14 0.0 is shown in fig.[fig1 ] .
a quasi - tetragonal cell with @xmath16 5.474(1 ) and @xmath17 20.88(1 ) was tentatively applied to the data , and a clear assignment of ( @xmath18 ) numbers to all the major peaks was obtained , indicating the high quality of the sample .
the one probable impurity phase ( less than 1% ) was srrho@xmath0 @xcite .
the major phase in the samples could be reasonably assigned to sr@xmath0rh@xmath1o@xmath2 , of which there are no known records in the literature . ) of sr@xmath0rh@xmath1o@xmath2 , obtained at room temperature . a tentatively applied quasi - tetragonal unit cell with @xmath16 5.474(1 )
and @xmath17 20.88(1 ) yields a clear assignment of ( @xmath18 ) numbers to almost all peaks , indicating quality of the sample .
unindexed peaks are marked by stars.,width=302 ] possible oxygen vacancies in the compound were quantitatively investigated in detail by thermogravimetric analysis .
approximately 30 milligrams of the @xmath14 0.0 powder was slowly heated at 5 @xmath10c / min in a gas mixture ( 3% hydrogen in argon ) and held at 800 @xmath10c until the weight reduction became saturated .
calculated oxygen composition from the weight loss data was 7.01 per formula unit , indicating good oxygen stoichiometry .
another sample of sr@xmath0rh@xmath1o@xmath2 , prepared at identical synthesis condition , was used for structural study by means of neutron diffraction .
the powder neutron data of powder sample of sr@xmath0rh@xmath1o@xmath2 ( @xmath190.2 grams ) were collected at room temperature using the bt-1 high - resolution powder diffractometer at the nist center for neutron research , employing a cu(311 ) monochromator to produce a monochromatic neutron beam of wavelength 1.5396 .
collimators with horizontal divergences of 15@xmath20 , 20@xmath20 , and 7@xmath20 of arc were used before and after the monochromator , and after the sample , respectively .
the intensities were measured in steps of 0.05@xmath10 in the 2@xmath21 range 3@xmath10168@xmath10 .
the structural parameters were refined using the program gsas @xcite .
the neutron scattering amplitudes used in the refinements were 0.702 , 0.593 , and 0.581 ( @xmath2210@xmath23 cm ) for sr , rh , and o , respectively .
the magnetic properties of the sample ( @xmath240.0 ) were studied in a commercial apparatus ( quantum design , mpms - xl system ) between 2 k and 390 k. the magnetic susceptibility data were collected at 1 koe and 70 koe , and magnetization curves were recorded between -70 koe and 70 koe after cooling the sample at each 2 k , 5 k and 50 k. the electrical resistivity was measured by a conventional ac - four - terminal technique at zero field and at 70 koe .
the ac - gauge current was 0.1 ma at 30 hz . the selected sample ( @xmath240.0 )
was cut out into a bar shape , and the each face was polished with aluminum - oxide - lapping film . in order to decrease the contact resistance with the sample , gold pads ( @xmath19 200 nm in thickness ) were deposited at four locations along the bar , and silver epoxy was used to fix platinum wires ( @xmath19 30 @xmath25m@xmath26 ) at each gold terminal .
the contact resistance at the four terminals was less than 8 ohm .
[ cols="<,<,<,<,<,<,<",options="header " , ] the crystal structure of the title compound was investigated by means of neutron powder diffraction at room temperature . in an effort to maximize the homogeneity of the sample ,
only one pellet was subjected to the neutron study rather than a mixture of multiple pellets , even though the total sample mass was far below the regular level .
the degree of local structure distortions , such as cooperative rotations of metal - oxygen octahedra and frequency of stacking faults in the layered structure , may depend sensitively on the synthesis conditions .
technical problems , such as small sample volume within the high - pressure apparatus and uncertainties in the maximum temperature during the heating process , could contribute to small variations in sample quality .
measuring an individual pellet by neutron powder diffraction was intended to preclude possible errors in the analysis of the local structural distortion .
the structural distortion was naturally expected in sr@xmath0rh@xmath1o@xmath2 based on the analogy with sr@xmath0ru@xmath1o@xmath2 , since the ionic size of @xmath27rh@xmath28 ( 0.60 ) @xcite is very close to that of @xmath27ru@xmath29 ( 0.62 ) @xcite .
furthermore , the expected distortions were hard to detect with normal x - ray diffraction .
the pellet as made in a platinum capsule was highly polished with a fine sandpaper to reduce possible pt contamination .
the pellet ( @xmath190.2 grams ) was then finely ground for the neutron diffraction study .
the neutron data were collected at room temperature for approximately one and a half days ; the profile is shown in fig.[fig2 ] .
although the sample powder was exposed to the neutron beam for many hours , the intensity did not reach normal levels due to the small sample mass .
it was , however , sufficient enough to deduce structural information from the data .
the small impurity contribution , found in the x - ray study , and possible magnetic contributions were not investigated due to the low intensity level .
0.2 grams ) of sr@xmath0rh@xmath1o@xmath2 , obtained at room temperature at @xmath30 1.5396 .
vertical bars indicate expected peak positions for the orthorhombic ( _ bbcb _ ) structure model .
the difference between the orthorhombic model ( solid lines ) and the data ( crosses ) is shown below the bars column .
although intensity statistic appeared to be rather lower due to the small sample mass , the analysis succeeded in obtaining a high - quality solution.,width=302 ] rh@xmath1o@xmath2 .
fine lines signifies the orthorhombic unit cell .
cooperative rotations along @xmath31axis of the rho@xmath32 octahedra are indicated by circular arrows.,width=302 ] the major concern in the structure investigation of the title compound was whether a cooperative rotation of the metal - oxygen octahedra occurs as is found in the analogous sr@xmath0ru@xmath1o@xmath2 @xcite . in the study of the @xmath33 oxide by means of neutron diffraction , eight independent modes for the cooperative rotation ( @xmath34 , @xmath35 , @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , @xmath41 ) , combinations of some of those , and the rotation - free mode ( @xmath42 ) were tested .
it was found that the @xmath35 mode resulted in the highest quality of the profile analysis @xcite . here ,
in the analysis of the profile of sr@xmath0rh@xmath1o@xmath2 , we first applied the @xmath35 mode and found it to produce reasonable solutions ( @xmath43 5.42% , @xmath44 6.47% ) to describe the cooperative rotation of rho@xmath32 octahedra in sr@xmath0rh@xmath1o@xmath2 .
although the @xmath35 mode is the most likely structural distortion in sr@xmath0ru@xmath1o@xmath2 , other lower symmetry modes , as mentioned above , are still possible .
unfortunately , the low intensity of the data in the present study made it difficult to completely dismiss the other modes .
detailed analysis of the issue is left for future work , in which further progress of the high - pressure - synthesis technique will allow us to obtain a larger amount of highly homogenized sample .
probable stacking faults were examined by introducing the o(3 ) elements with some constraints ( caption of table [ table1 ] ) , as was done for sr@xmath0ru@xmath1o@xmath2 . as a result ,
significant improvements in the overall fit were obtained ( @xmath452% in @xmath46 values ) .
the estimated occupancy of o(3 ) was @xmath198% , indicating the frequency of the stacking faults .
the estimated structural parameters of sr@xmath0rh@xmath1o@xmath2 , and selected bond distance and angles , are arranged in table [ table1 ] .
the isotropic atomic displacement parameter ( @xmath47 ) of the rhodium ion ( 8@xmath48site ) is rather unusual , possibly because the low level of the intensity statistics reduced reliability of the thermal parameters somewhat .
we believe further structure studies , on not only a larger amount of sample powder , but also a single crystal of sr@xmath0rh@xmath1o@xmath2 , might provide much more reliable data .
the structure of sr@xmath0rh@xmath1o@xmath2 is shown in fig.[fig3 ] .
the directions of rotation are indicated by circular arrows .
the rotation results in the bond angle of rh
o(3)rh being 159 degrees rather than 180 degrees for an ideally rotation - free structure .
the rotation angle of rho@xmath32 octahedra along @xmath31axis is 10.5 degrees , which is slightly greater than that of ruo@xmath32 octahedra in the analogous sr@xmath0ru@xmath1o@xmath2 ( 6.8 degrees ) @xcite .
a general diagram for the degree of distortion in ruddlesden - popper - type compounds @xmath49o@xmath2 ( @xmath50 alkaline and/or rare earth metal ; @xmath51 transition metal ) was proposed , in which the real members of @xmath49o@xmath2 were divided into two categories : one with and one without distortions associated with cooperative rotations of the @xmath52o@xmath32 octahedra @xcite . a criterion named `` zero strain line ''
was established for judging which category each member should belong to .
the combination of radii of @xmath27rh@xmath53 ( 0.60 ) and @xmath54sr@xmath55 ( 1.44 ) in sr@xmath0rh@xmath1o@xmath2 @xcite yields a point below the `` zero strain line '' in the proposed diagram , implying that sr@xmath0rh@xmath1o@xmath2 should be in the distortion - free category .
this classification for the sr@xmath0rh@xmath1o@xmath2 case appears to be incorrect as dictated by the data in the present study @xcite .
characteristics of chemical bonds may play a pivotal role in the distortion process , and their consideration may be necessary if a complete diagram is to be developed .
rh@xmath1o@xmath2 at zero filed and 70 koe .
inset shows an expansion of the low - temperature portion.,width=302 ] the electrical resistivity data of the polycrystalline sample of sr@xmath0rh@xmath1o@xmath2 are shown in fig.[fig4 ] .
temperature and magnetic field dependence was studied between 2 k and 390 k. the resistivity at room temperature is @xmath1975 m@xmath56-cm , which is approximately one magnitude lager than that of the 3d srrho@xmath0 @xcite and one magnitude smaller than that of the 2d sr@xmath1rho@xmath4 , which is a barely metallic conductor @xcite .
the series of the rhodium oxides sr@xmath57rh@xmath58o@xmath59 ( @xmath60 1 , 2 , and @xmath61 ) consist of solely homovalent elements including rh@xmath53 ( 4@xmath62 may be in the low - spin configuration @xmath63 ) , and all of them clearly show metallic character , while the electronic dimensionality varies from two to three .
this suggests that the metallic character is developed with increasing electronic dimensionality .
further studies involving band structure calculations could be significant in revealing how essential the dimensionality is to the electronic transport of the rhodium oxide series .
a small magnetic field dependence was found at low temperature ( @xmath64k ) as shown in the inset of fig.[fig4 ] .
future studies on single crystals may provide insight into the mechanism responsible for the positive magnetoresistance .
rh@xmath1o@xmath2 and srrho@xmath0 .
the data for the perovskite compound were taken from a previous report @xcite.,width=302 ] the seebeck coefficient of the title compound was measured and compared with that of the perovskite srrho@xmath0 ( fig.[fig5 ] ) @xcite .
the seebeck coefficient of sr@xmath1rho@xmath4 , reported previously , was approximately @xmath65v / k at room temperature and decreases linearly to approach zero upon cooling @xcite .
the data of all three compounds ( sr@xmath1rho@xmath4 , sr@xmath0rh@xmath1o@xmath2 , and srrho@xmath0 ) are qualitatively consistent with the metallic conductivity observed for each , and indicate that the sign of the majority carriers is positive in the metallic conducting state . among the seebeck coefficient of the three compounds , it is hard to find a remarkable difference qualitatively and quantitatively , in contrast to the case of the resistivity data .
although a small change in slop in the seebeck coefficient was observed in both sr@xmath0rh@xmath1o@xmath2 and srrho@xmath0 at approximately 150 k ( fig.[fig5 ] ) , there were , however , no corresponding anomalies in other transport data around that temperature . the change may be structurally related , and further studies of the temperature dependence of structural distortions may help to address the issue .
the data are plotted in @xmath66 vs @xmath67 ( main panel ) and @xmath68 vs @xmath69 ( inset ) forms .
the dotted curve represents a fit to the data ( details are in the text).,width=302 ] the specific heat data are plotted as @xmath66 vs @xmath67 ( main panel ) and @xmath68 vs @xmath69 ( inset ) in fig.[fig6 ] .
the electronic and lattice contributions to the specific heat can be isolated from the data in the main panel by using the following low - temperature form ( @xmath70 ) : @xmath71 where @xmath72 , @xmath73 , @xmath74 , @xmath75 , and @xmath76 are debye temperature , sommerfeld s coefficient , boltzmann s constant , and avogadro s number , respectively .
the two independent parameters , @xmath72 and @xmath73 , which indicate physical characteristics of the compound , were estimated by a least - square method in the linear range between 150 k@xmath77 and 450 k@xmath77 ; we obtained @xmath78 172 k and @xmath79 -0.00199 j@xmath80mol - rh@xmath81k@xmath82 .
the estimation of both @xmath73 and @xmath72 was , however , not solid ; a small degree of variations was found .
for example , when a @xmath83 term was added to the equation , the fits achieved were @xmath84 of @xmath19 180 k , @xmath73 of @xmath190.00365 j@xmath80mol - rh@xmath81k@xmath85 , and @xmath74 ( the coefficient of the @xmath83 term ) of 7.93@xmath2210@xmath86 j@xmath80mol - rh@xmath81k@xmath87 , as indicated by the dotted curve .
although it is technically difficult to determine a precise value for @xmath73 , it should , however , be very small in contrast to that of the metallic srrho@xmath0 ( @xmath73 = 7.6 mj / mol - rh / k@xmath77 ) @xcite .
the depressed metallic character of the bilayered sr@xmath0rh@xmath1o@xmath2 is reflective in the very small value of @xmath73
. it would be interesting to compare the value of @xmath88 , computed from the density of states at the fermi level by means of band - structure calculations , with and without consideration of spin polarization . in the above analysis we assumed that @xmath89 was approximately equal to @xmath90 because it was in the low temperature limit . as expected from the debye temperature ( 170 k180 k ) of the layered oxide sr@xmath0rh@xmath1o@xmath2 , even within the studied temperature range , the specific heat approaches the high - temperature limit , roughly calculated to be @xmath19150 j@xmath80mol - rh@xmath81k@xmath91 [ 6(atoms per unit cell ) @xmath223 ( dimensionality per atom ) @xmath92n(boltzmann and avogadro s constants ) ] .
the debye temperature of sr@xmath0rh@xmath1o@xmath2 is close to that of the perovskite srrho@xmath0 ( @xmath78 190 k ) . to our knowledge
, the debye temperature of sr@xmath1rho@xmath4 has not been reported .
a small component , which appears in the low - temperature limit of the @xmath93 vs @xmath67 curve ( the data points gradually apart from the fitting curve on cooling ) , is probably due to a magnetic origin . in order to make clear the origin of the probable magnetic component
, several magnetic models were tested on the specific heat data .
the studies , however , did not yield a clear assignment of magnetic model to the origin because there were still too many possibilities related to a nuclear schottky specific heat , an empirical temperature - independent term , ferro- and antiferromagnetic spin fluctuations , and associations of some of those .
as far as the present investigation goes , it was therefore unlikely to establish a clear picture with a reasonable sense to understand the appearance of the very small component in the low - temperature specific heat data .
rh@xmath1o@xmath2 at 1 koe and 70 koe on cooling , and applied magnetic field dependence of the magnetization ( inset ) at 2 k , 5 k and 50 k.,width=302 ] nearly - temperature - independent magnetic susceptibility data , measured between 2 k and 390 k , are shown in fig.[fig7 ] .
the majority of the data could be characterized as pauli paramagnetic with an enhanced @xmath94 .
the @xmath95 is approximately 1.0@xmath96 emu / mol - rh , nearly one magnitude larger than those of normal paramagnetic metals , including rh(iv)o@xmath1 @xcite .
this is a curious result , considering the very small @xmath73-value estimated from the specific heat measurements .
more will be said on this below .
alternatively , a layered spin system , despite the metallic character , is probable to account for the magnetic characteristics with the broad maximum at about 200 k as is the case for the sr doped la@xmath1cuo@xmath4 @xcite .
further studies to explore probable localized magnetic moments should be of interest .
the steep upturn in low temperature at 1 koe is due to a small concentration of magnetic impurities , which could be well fit to a curie - weiss law ; the magnetic parameters were @xmath97 0.00194 emu k / mol - rh ( curie constant ) , @xmath98 -0.1462 k ( weiss temperature ) , and @xmath99 9.21@xmath100 emu / mol - rh ( temperature - independent susceptibility ) . the rather small value of @xmath101 , less than 1 k , indicates that the diluted moments are free from magnetic interactions , as is often the cases .
the @xmath52 vs @xmath102 curves measured at 2 k , 5 k and 50 k are shown in the inset of fig.[fig7 ] . the small component ( @xmath190.001@xmath103 at 70 koe and 5 k ) , which is superimposed on the paramagnetic background , is probably due to the magnetic impurity outlined above .
while a metamagnetic transition was clearly found in the analogous sr@xmath0ru@xmath1o@xmath2 at @xmath19 50 koe and 2.8 k in fields both parallel and perpendicular to the @xmath31axis of a single crystal @xcite , no trace of one could be seen in the data of sr@xmath0rh@xmath1o@xmath2 even at 2.0 k. such a transition may be shifted to higher fields in the rh sample , and in order to investigate this possibility , a pulsed high - magnetic - field study at low temperature is scheduled for sr@xmath0rh@xmath1o@xmath2 .
high - pressure synthesis techniques have been utilized in an effort to add to the variety of experimental systems currently being used in the study of quantum phase transitions and criticality .
the systematic synthesis experiments performed thus far on the rh@xmath53 ( 4@xmath104 , expected ) oxide system have revealed a new class of ruddlesden - popper phases sr@xmath57rh@xmath105o@xmath59 ( @xmath60 1 [ ref . 31 and 32 ] , 2
[ this work ] , and @xmath61 [ ref .
the neutron diffraction study clearly reveals that the new metallic rhodium oxide , sr@xmath0rh@xmath1o@xmath2 , is isostructural to the ruthenium oxide sr@xmath0ru@xmath1o@xmath2 , which has received much recent attention due to its intriguing quantum characteristics .
the magnetic and transport data presented here , however , do not indicate clear contributions from quantum fluctuations in the studied range of temperature and magnetic field , in contrast to what was observed in the ruthenium compound @xcite . since the curie - weiss - like behavior was not seen ( except for the impurity term ) even in the high - temperature portion of the magnetic susceptibility data , the self - consistent renormalization theory of spin fluctuations for both antiferro- and ferromagnetic nearly - ordered magnetic metals may not be applicable to the present magnetic and specific heat data . while the magnetic susceptibility data of sr@xmath0rh@xmath1o@xmath2 suggest a rather large density of states at the fermi level ( similar to what was observed in srrho@xmath0 ) , the linear term in the heat capacity is , however , remarkably small on the order of semimetals like bi and sb .
the resistivity is about 10 times larger than that of srrho@xmath0 , qualitatively consistent with the rather small @xmath73 .
it is possible that the influence of disorder in the polycrystalline sample at a level undetectable by the techniques in the present study masks the intrinsic nature of ideally clean sr@xmath0rh@xmath1o@xmath2 . in the studies of the ruthenium oxides , for example
, the quantum properties were found to be rather fragile to such disorder @xcite .
further information , including carrier density , degree of disorder , band structure , would prove useful in clearly identifying the intrinsic properties of sr@xmath0rh@xmath1o@xmath2 .
high - quality single crystals of sufficient size for electronic transport studies synthesized in the high - pressure cell would afford the opportunity of further investigations of sr@xmath0rh@xmath1o@xmath2 at a much deeper level . in the ruthenium oxide family , much effort
have been spent in searching for new superconducting compounds , but only the 214 phase is a superconductor .
one member found recently ( non - superconducting ) is the double - layered sr@xmath0ru@xmath1o@xmath2f@xmath1 , where ru ions are pentavalent @xcite .
one electronic hole per formula unit was formally doped into metallic sr@xmath0ru@xmath1o@xmath2 , and an ordered magnetic moment ( @xmath191@xmath106 per ru ) was found below about 185 k. on the other hand , sr@xmath0rh@xmath1o@xmath2 has five 4@xmath3 electrons per formula unit , i.e. one electron per ru is hypothetically doped into sr@xmath0ru@xmath1o@xmath2 , and no localized moments observed .
the following is a general description of the three bilayered compounds : the localized character at the 4@xmath107 layered compound gets weakened at 4@xmath108 and becomes much more itinerant at 4@xmath109 .
the overall comprehensive picture about the electronic structure of various 4@xmath3 compounds has probably not yet been established .
further studies , including theoretical considerations and a systematic study of isovalent and aliovalent chemical doping to the series sr@xmath57rh@xmath58o@xmath110 would be of interest .
further progress of the high - pressure - synthesis technique may allow future investigations of new and unforeseen electronic materials those are potentially intriguing . we wish to thank dr .
m. akaishi ( nims ) and dr .
s. yamaoka ( nims ) for their advice on the high - pressure experiments .
this research was supported in part by superconducting materials research project , administrated by the ministry of education , culture , sports , science and technology of japan .
one of us ( k.y . ) was supported by the domestic research fellowship , administrated by the japan society for the promotion of science .
the words `` clean metal '' hereby mean a metal in the low disorder limit .
stewart , rev .
* 73 * , 797 ( 2001 ) .
s. sachdev , _ quantum phase transitions _ , cambridge university press , cambridge , 1999 .
sondhi , s.m .
girvin , j.p .
carini , and d. shahar , rev .
phys . * 69 * , 315 ( 1997 ) .
j.a . hertz , phy .
b * 14 * , 1165 ( 1976 ) .
k. izawa , h. takahashi , h. yamaguchi , yuji matsuda , m. suzuki , t. sasaki , t. fukase , y. yoshida , r. settai , and y. onuki , phy .
lett . * 86 * , 2653 ( 2001 ) .
k. ishida , y. kitaoka , k. asayama , s. ikeda , s. nishizaki , y. maeno , k. yoshida , and t. fujita , phy .
b * 56 * , r505 ( 1997 ) .
grigera , r.s .
perry , a.j .
schofield , m. chiao , s.r .
julian , g.g .
lonzarich , s.i .
ikeda , y. maeno , a.j .
millis , and a.p .
mackenzie , science * 294 * , 329 ( 2001 ) .
perry , l.m .
galvin , s.a .
grigera , l. capogna , a.j .
schofield , a.p .
mackenzie , m. chiao , s.r .
julian , s.i .
ikeda , s. nakatsuji , y. maeno , and c. pfleiderer , phy .
lett . * 86 * , 2661 ( 2001 ) .
singh and i.i .
mazin , phy .
b * 63 * , 165101 ( 2001 ) . s.i . ikeda , y. maeno , s. nakatsuji , m. kosaka , and y. uwatoko , phy .
b * 62 * , r6089 ( 2000 ) .
puchkov , m.c .
schabel , d.n .
basov , t. startseva , g. cao , t. timusk , and z .- x .
shen , phy .
81 * , 2747 ( 1998 ) .
k. yoshimura , t. imai , t. kiyama , k.r .
thurber , a.w .
hunt , and k. kosuge , phys .
lett . * 83 * , 4397 ( 1999 ) .
t. kiyama , k. yoshimura , k. kosuge , h. mitamura , and t. goto , j. phys .
. jpn . * 68 * , 3372 ( 1999 ) .
t. kiyama , k. yoshimura , k. kosuge , h. michor , and g. hilscher , j. phys .
. jpn . * 67 * , 307 ( 1998 ) .
g. cao , s. mccall , m. shepard , j.e .
crow , and r.p .
guertin , phys .
b * 56 * , 321 ( 1997 ) . t. he and r.j .
cava , phys .
b * 63 * , 172403 ( 2001 ) .
t. he , q. huang , and r.j .
cava , phys .
b * 63 * , 024402 ( 2000 ) .
k. yamaura and e. takayama - muromachi , phys .
b * 64 * , 224424 ( 2001 ) .
stitzer , a.el abed , j. darriet , and h .- c .
zur loye , j. am .
soc . * 123 * , 8790 ( 2001 ) .
plaisier , a.a.c .
van vliet , and d.j.w .
ijdo , j. alloys comp . * 314 * , 56 ( 2000 ) .
j.b . claridge and h .- c .
zur loye , chem . mater . * 10 * , 2320 ( 1998 ) .
r. hory , z. bukowski , m. wocyrz , and a.j .
zaleski , j. alloys comp . *
262 - 263 * , 267 ( 1997 ) .
t. nakamura , t. shimura m. itoh , and y. takeda , j. sold state chem . * 103 * , 523 ( 1993 ) .
s. yamaoka , m. akaishi , h. kanda , t. osawa , t. taniguchi , h. sei , and o. fukunaga , j. high pressure inst . jpn . * 30 * , 249 ( 1992 ) .
a.c . larson and r.b .
von dreele , los alamos national laboratory report no .
laur086 - 748 ( 1990 ) .
r. d. shannon , acta crystallogr .
a 32 , 751 ( 1976 ) .
h. shaked , j.d .
jorgensen , o. chmaissem , s. ikeda , and y. maeno , j. solid state chem . * 154 * , 361 ( 2000 ) .
h. shaked , j.d .
jorgensen , s. short , o. chmaissem , s .- i .
ikeda , and y. maeno , phy .
b * 62 * , 8725 ( 2000 ) .
q. huang , j.w .
lynn , r.w .
erwin , j. jarupatrakorn , and r.j .
cava , phy .
b * 58 * , 8515 ( 1998 ) .
m. itoh , t. shimura , y. inaguma , and y. morii , j. solid state chem .
* 118 * , 206 ( 2000 ) .
t. shimura , m. itoh , y. inaguma , and t. nakamura , phy .
b * 49 * , 5591 ( 1994 ) .
k. yamaura , d. p. young , and e. takayama - muromachi , in the 2002 mrs spring meeting , san francisco , california ( in press ) .
h. kojima , r.s .
tebble , and d.e.g .
williams , proc .
london * a260 * , 237 ( 1961 ) .
budworth , f.e .
hoare , and j. preston , proc .
london * a257 * , 250 ( 1960 ) .
hoare and j.c .
walling , proc .
london , sect .
b * 64 * , 337 ( 1951 ) .
m. imada , a. fujimori , and y. tokura , rev .
phys . * 70 * , 1039 ( 1998 ) .
s. ikeda , y. maeno , and t. fujita , phy .
b * 57 * , 978 ( 1998 ) .
cava , h.w .
zandbergen , j.j .
krajewski , w.f .
, b. batlogg , s. carter , r.m .
fleming , o. zhou , and l.w .
rupp jr . , j. solid state chem . * 116 * , 141 ( 1995 ) .
t. moriya , _ spin fluctuations in itinerant electron magnetism _ , edited by manuel cardona ( springer - verlag , 1985 )
. t. moriya and k. ueda , adv .
phys . * 49 * , 555 ( 2000 ) .
l. capogna , a.p .
mackenzie , r.s .
perry , s.a .
grigera , l.m .
galvin , p.raychaudhuri , a.j .
schofield , c.s .
alexander , g. cao , s.r .
julian , and y. maeno , phy .
lett . * 88 * , 076602 ( 2002 ) .
mackenzie , r.k.w .
haselwimmer , a.w .
tyler , g.g .
lonzarich , y. mori , s. nishizaki , and y. maeno , phy .
80 * , 161 ( 1998 ) .
li and c. greaves , phy .
b * 62 * , 3811 ( 2000 ) . | the bilayered rhodium oxide sr@xmath0rh@xmath1o@xmath2 was synthesized by high - pressure and high - temperature heating techniques .
the single - phase polycrystalline sample of sr@xmath0rh@xmath1o@xmath2 was characterized by measurements of magnetic susceptibility , electrical resistivity , specific heat , and thermopower .
the structural characteristics were investigated by powder neutron diffraction study .
the rhodium oxide sr@xmath0rh@xmath1o@xmath2 [ _ bbcb _ , _ a _
= 5.4744(8 ) , _ b _ = 5.4716(9 ) , _ c _
= 20.875(2 ) ] is isostructural to the metamagnetic metal sr@xmath0ru@xmath1o@xmath2 , with five 4@xmath3 electrons per rh , which is electronically equivalent to the hypothetic bilayered ruthenium oxide , where one electron per ru is doped into the ru-327 unit .
the present data show the rhodium oxide sr@xmath0rh@xmath1o@xmath2 to be metallic with enhanced paramagnetism , similar to sr@xmath0ru@xmath1o@xmath2 .
however , neither manifest contributions from spin fluctuations nor any traces of a metamagnetic transition were found within the studied range from 2 k to 390 k below 70 koe . |
a radical pair ( rp ) is a pair of transient radicals whose unpaired electron spins , one on each radical , are correlated .
provided the rp is formed in a spin conserving process , such as photo - induced electron transfer in a precursor molecule , the initial spin state is known and this information evolves over time under the influence of external magnetic fields ( zeeman interactions ) as well as internal electron - electron ( exchange and dipolar ) and electron - nuclear hyperfine coupling . in the simplest reaction scheme , the rp recombines to different products depending on the electronic spin state
singlet , s or triplet , t ( fig . 1 ) .
the relative yields of the two products therefore contain information about the interactions experienced by the rp during its lifetime , including a dependence on the direction of the external magnetic field .
this is the physical basis of the proposal that rp reactions are responsible for the magnetic compass sense of migratory birds.@xcite there is a growing body of evidence in support of this proposal .
behavioural studies have shown that migratory birds are only magnetically oriented in the presence of blue / green light.@xcite this is broadly consistent with the photo - excitation of the flavin adenine dinucleotide ( fad ) cofactors in cryptochrome proteins in the avian retina which are believed to be the precursors of the rp reactions.@xcite studies have also shown that the avian magnetic sensor is an inclination compass rather than a polarity compass : it is insensitive to inverting the direction of the magnetic field.@xcite this is consistent with time - reversal symmetry in the rp mechanism . and more recent studies have found that birds are disoriented when exposed to weak radiofrequency ( rf ) electromagnetic fields.@xcite if this were properly understood , it might perhaps provide the most compelling evidence of all for rp - based magnetoreception.@xcite however , theoretical calculations of the effect of weak rf magnetic fields on the rp mechanism are currently limited , and the existing experimental evidence is still somewhat controversial . the effect of rf radiation on avian magnetoreception was first observed in a study by ritz _ et al._@xcite in 2004 , in which both broadband noise and single frequency electromagnetic fields were found to disrupt the ability of birds to magnetically orient . following this initial finding ,
there have been a number of further experimental studies aimed at characterising the interaction between the rf magnetic field and the magnetosensor.@xcite one of the most influential of these found that irradiation at the larmor " frequency the precession frequency of a free electron spin in the earth s magnetic field ( 1.4 mhz at 50 @xmath0 t ) had a special resonance effect , causing disorientation at even very weak intensity levels ( @xmath1 nt).@xcite if this were correct , it would imply that the electron spin on one of the radicals in the rp magnetosensor is coupled only to the external field , and not to any nuclear spins .
this in turn would impose a stringent constraint on the identity of the rp , as there are few radicals without any nuclear spins in biological systems .
however , a replication study has since failed to show the same sensitivity to larmor frequency perturbations , and found instead that broadband noise is far more effective at disrupting the avian compass.@xcite given these conflicting experimental results , a theoretical study of how the time - dependent magnetic field component of rf radiation affects the rp spin dynamics would clearly be very useful .
however , there are several obstacles to such a study .
firstly , at least one ( and quite possibly both ) of the biological radicals involved is likely to have many nuclear spins with anisotropic hyperfine interactions . since the size of the hilbert space increases exponentially with the number of spins , even simulations involving a static external field are computationally expensive .
secondly , if we wish to treat broadband noise perturbations , the time dependence of the external field will not be monochromatic as is required for many existing approximate methods.@xcite additionally , as noted by gauger _
et al._,@xcite for a rp to exhibit the observed sensitivity to weak time - dependent fields , its lifetime must be exceedingly long .
one exact method that is applicable to the problem is the numerical solution of the liouville - von neumann equation , but this becomes cripplingly computationally expensive for long simulation times and realistic rps .
an alternative is provided by the compute algorithm,@xcite which addresses the need for a long simulation time by constructing the propagator for a full modulation period of the ( periodic ) rf field .
we have considered using this algorithm for the present problem , but found that a large number of matrix exponentials are needed to construct the propagator when the rf field contains many fourier components , and that the expense of evaluating these exponentials becomes prohibitive when the rp contains many nuclear spins .
we therefore require a new approximate method capable of treating large spin systems in the presence of weak external magnetic fields with a complex time dependence ( involving many fourier components ) .
in the following sections , we shall first present and then validate such a method , based on a modified formulation of floquet theory .
various versions of this theory are well known and widely used in fields ranging from solid state nuclear magnetic resonance to multiphoton spectroscopy.@xcite however , we are not aware of any previous application of floquet theory to radical pair recombination reactions in time - dependent magnetic fields .
$ ] and @xmath2 $ ] interconvert by hyperfine- and zeeman - mediated intersystem crossing , and recombine to give distinct products with rate constants @xmath3 and @xmath4 . ]
we are interested in treating a quantum system evolving under a hamiltonian which is periodic in time with period @xmath5 , @xmath6 shirley@xcite was the first to employ floquet s theorem@xcite to show that this is equivalent to solving an infinite dimensional eigenvalue problem involving a so - called floquet hamiltonian with matrix elements @xmath7 here @xmath8 is a basis state in an extended floquet space " , the product of a hilbert space basis state @xmath9 and a fourier space basis state @xmath10 with @xmath11 , and @xmath12 in particular , shirley showed that the matrix element of the evolution operator between states @xmath9 and @xmath13 in hilbert space can be written as@xcite @xmath14 the eigenvalues @xmath15 and eigenstates @xmath16 of the floquet hamiltonian exhibit certain periodicity properties , which are endowed by the structure of @xmath17,@xcite @xmath18 @xmath19 these relations are in fact required to ensure the unitarity of the propagator in eq .
( 4 ) , and they can be used to re - write @xmath20 as follows:@xcite @xmath21 consider the expectation value of an observable @xmath22 at time @xmath23 subject to some initial condition at time @xmath24 . in the density operator formulation , this is @xmath25,\ ] ] where @xmath26 is the density operator at time @xmath24 . inserting resolutions of the identity in hilbert space ,
this becomes @xmath27 where @xmath28 and @xmath29 .
if we use eq .
( 7 ) for the matrix elements of the time evolution operator , we obtain @xmath30 introducing a trace over floquet space , this can be written in a form analogous to eq .
( 8) , @xmath31,\ ] ] where the floquet space detection and density operators are @xmath32 @xmath33 now had we used eq .
( 4 ) for @xmath20 rather than eq . ( 7 ) , we would have obtained the formulation of floquet theory described by ernst and co - workers,@xcite in which the floquet space detection and density operators are @xmath34 @xmath35 however , this is less convenient for our purposes , because the detection operator @xmath36 contains time - dependent phase factors .
these lead to less convenient time - independent expressions for observables such as the singlet yield of a radical pair recombination reaction that we shall consider below . our initial density operator @xmath37 also contains phase factors , but these are easier to deal with , because they are evaluated at time @xmath24 rather than time @xmath23 .
indeed it is often natural to specify the initial conditions at time @xmath38 , in which case the phase factors in @xmath37 disappear and eqs .
( 11 ) and ( 13 ) simplify to @xmath39,\ ] ] @xmath40 if the eigenspectrum of @xmath41 can be found or approximated , eq . ( 16 ) can be evaluated by inserting resolutions of the identity in floquet space , @xmath42 before we move on to discuss the diagonalisation of @xmath41 , it might be useful to summarise the various operators that appear in eq .
( 16 ) in matrix notation : @xmath43 @xmath44 @xmath45 here each individual block ( separated by horizontal and vertical lines ) is the size of the hilbert space , and @xmath46 is the hilbert space identity matrix .
the evaluation of eq . ( 18 ) requires the solution of a _ time - independent _ matrix problem : the diagonalisation of @xmath47 and the transformation of @xmath48 and @xmath49 into its eigenstate basis . clearly ,
because the matrices involved are infinite dimensional , this can only be done approximately . one way to tackle
the problem is to truncate the floquet hamiltonian , including a certain number of hilbert space sized blocks and diagonalising this enlarged matrix numerically.@xcite this method may offer a practical solution when there are only one or two fourier components in the hamiltonian , but if we wish to extend the method to include thousands of values of @xmath50 and a high - dimensional hilbert space , truncating @xmath47 is no longer viable .
fortunately , for the problem we shall be interested in ( a radical pair in a weak rf magnetic field ) , the time - dependent interactions are orders of magnitude smaller than the static hamiltonian @xmath51 , so we can use a perturbative method for the diagonalisation .
consider for simplicity the case of a single applied fourier component @xmath52 this form of the hamiltonian makes @xmath47 block tri - diagonal , with @xmath53 in the super- and sub - diagonal blocks .
if the hilbert space basis is the eigenbasis of @xmath51 , it is clear that there will be a degeneracy between a state in the @xmath54 diagonal block of @xmath47 and another in the @xmath55 block if the applied frequency @xmath56 is resonant with the energy difference between two unperturbed eigenvalues .
if the oscillating field is sufficiently weak , its effect will be confined to these resonant interactions .
we shall therefore assume this to be the case , and use degenerate perturbation theory to calculate the first order corrections to the states brought into resonance ( or near - resonance ) by @xmath57 . as noted by shirley,@xcite a near - resonance between two states , @xmath58 ,
means there are nearly - degenerate two - dimensional sub - spaces involving state @xmath59 in diagonal block @xmath50 and state @xmath60 in block @xmath61 .
the periodicity of @xmath47 implies that there is a copy of the same @xmath62 sub - matrix between all pairs of adjacent diagonal blocks @xmath63 shifted by an integer multiple of @xmath64 .
+ diagonalising this @xmath65 sub - matrix with the unitary matrix @xmath66 gives the first order corrected energies for states @xmath60 and @xmath59 as the eigenvalues @xmath67 and @xmath68 . since the multiple of @xmath64 simply shifts eigenvalues , the altered energies and eigenstates can be written as @xmath69 these approximate eigenvalues and eigenstates satisfy the periodicity conditions in eqs .
( 5 ) and ( 6 ) , as required .
the other ( non - resonant ) energies and eigenstates are unaltered , @xmath70 and @xmath71 , and we now have an approximate eigendecomposition of @xmath47 to first order in @xmath72 .
clearly , this approximation rests on the assumption that @xmath73 , because if this were not the case the @xmath62 matrix in eq .
( 23 ) would not be isolated from other interactions .
the method described above can be extended to treat more complicated situations including degeneracies in the unperturbed system , overlapping resonances , and a larger number of fourier components , provided again that the spacing @xmath56 between these fourier components is large compared with the strength of the perturbation .
the general procedure is to build and diagonalise sub - matrices including as many as possible of the resonances and the appropriate coupling terms to obtain the first order energies .
if there are multiple nearly - degenerate states , the dimension of the corresponding sub - space is enlarged to include them
. it should be apparent that each state can only appear in one nearly - degenerate sub - matrix , all of the states with which is it nearly degenerate being by extension close in energy to one another .
one can construct examples to show that it may not always be possible to include _ all _ near resonances without allowing the size of the nearly - degenerate sub - matrix to exceed the size of the hilbert space .
however , if the hilbert space contains @xmath74 states , it is always possible to build an @xmath75 nearly - degenerate sub - matrix of @xmath47 that contains one copy of each hilbert space state and captures ( at least ) the @xmath76 _ closest _ resonances ( and typically many more ) .
assuming that the off - diagonal coupling terms in the floquet hamiltonian all have similar orders of magnitude , it is these closest resonances that will have the largest effect on the perturbed eigenvalues and eigenstates .
neglecting more distant near resonances is clearly an approximation , but it is a convenient one to make , because diagonalising a hilbert space - sized sub - matrix of @xmath47 is no more expensive than diagonalising @xmath77 . since this procedure is also easily automated , it is the one we have adopted in the example calculations reported in sec .
v. having found an approximate eigendecomposition of @xmath47 , the final stage is to evaluate the expression for @xmath78 in eq .
( 18 ) involving a double sum over the floquet eigenstates .
the number of these eigenstates is infinite , but only a finite number of them contribute to the double sum because of the structure of @xmath79 in eq .
for example , in the simple case where the approximate floquet eigenstates are those in eq .
( 25 ) , the matrix element @xmath80 in eq . ( 18 ) is only non - zero when @xmath81 is 0 or 1 and @xmath50 is 0 or @xmath82 .
the infinite sums over @xmath50 and @xmath81 therefore each collapse to just two terms , and an analogous simplification is obtained in the more general case in which a hilbert space - sized nearly - degenerate sub - matrix of @xmath47 is diagonalised to obtain the approximate floquet eigenstates .
with this floquet machinery in hand , let us now return to the problem of simulating the spin dynamics of a rp subject to a rf magnetic perturbation . in the absence of any dipolar or exchange coupling between the two electrons , the unperturbed hamiltonian that governs the spin evolution of a rp in the presence of a static external zeeman field is@xcite @xmath83 where the individual radical hamiltonians are @xmath84 here the first term , in which @xmath85 is the gyromagnetic ratio of the electron in radical @xmath86 , is the zeeman interaction of the electron spin @xmath87 with the static magnetic field @xmath88 .
the second term contains the anisotropic hyperfine interactions between the electron spin and the @xmath89 nuclear spins @xmath90 in the radical .
we have neglected the comparatively weak zeeman interactions of the nuclear spins with the magnetic field and the nuclear spin - spin coupling , but these could easily be added to eq . (
27 ) if necessary . in the presence of a radiofrequency magnetic field , the full hamiltonian @xmath91 that governs the evolution of the radical pair has the form in eq .
( 1 ) , where @xmath51 is given in eq . ( 26 ) and when @xmath92 @xmath93 with @xmath94 here @xmath95 is the ( real ) magnetic field vector of the @xmath96 fourier component of the rf radiation , and @xmath97 is a phase factor analogous to the @xmath98 in eq .
note that the hermicity of @xmath91 demands that @xmath99 , which implies that @xmath100 and @xmath101 .
the rp will typically be produced by a photo - induced electron transfer reaction from a singlet ground electronic state with equilibrium nuclear spin states . since spin is conserved in this process , the initial density operator is@xcite @xmath102 where @xmath103 is the number of nuclear spin states in radical @xmath86 and @xmath104 is the singlet projection operator @xmath105 although there are other detection operators one could consider , the most relevant for our purposes is simply @xmath106 the corresponding time - dependent observable @xmath78 is the probability @xmath107 that the rp is still in the singlet state at time @xmath23 after the initial photo - excitation . according to eq . (
18 ) , this is given by @xmath108 where from eqs .
( 12 ) and ( 32 ) , @xmath109 and from eqs .
( 17 ) and ( 30 ) , @xmath110 note that eq .
( 33 ) ignores the possibility of electron spin relaxation , which would be difficult to include in the present formulation .
the effect of electron spin relaxation on the fad - tryptophan radical pair in cryptochrome has recently been investigated in a separate publication.@xcite figure 2 shows the singlet probability as a function of time for a model rp ( details given in sec .
v ) with and without a single @xmath111 nt resonant rf perturbation applied .
the plot highlights the oscillatory nature of @xmath107 , and also confirms the assertion of gauger _ _ et al.__@xcite that a very long ( @xmath112 ) rp spin coherence is required for there to be a significant rf field effect with such a weak perturbation .
the fact that this is many orders of magnitude longer than the period of the rapid oscillations in @xmath107 will be used to simplify the calculation of the singlet yield of the reaction below . in the context of avian magnetoreception ,
the observable of interest is the dependence of the singlet ( or equivalently , since they sum to one , the triplet ) yield of the radical pair recombination reaction on the direction of the static magnetic field.@xcite assuming for simplicity that the recombination is symmetric ( @xmath113 in fig . 1 )
, this singlet yield is given by@xcite @xmath114 inserting the expression for @xmath107 in eq .
( 33 ) and doing the integral over @xmath23 gives @xmath115 the form of this last equation reveals a simplification that is crucial for dealing with the infinite floquet space
. the factor of @xmath116 $ ] suppresses contributions to the double sum in eq .
( 37 ) whenever the energy difference @xmath117 is much larger than the recombination rate constant @xmath118 : the coherent oscillations that are fast on the timescale of the rp lifetime have a negligible effect on the singlet yield .
this implies that we can discard off - diagonal terms between states that differ widely in energy . in the case of long - lived radical pairs
, the majority of terms in the double sum can be discarded . indeed ,
when @xmath118 is sufficiently small , the only terms that contribute come from nearly - degenerate sub - spaces of the floquet space exactly the same sub - spaces that we encountered when diagonalising @xmath47 by degenerate perturbation theory in sec .
some manipulation of the floquet space detection operator will help us to exploit this simplification in the evaluation of @xmath119 . using the periodicity properties of the eigenvalues and eigenstates of @xmath47 in eqs .
( 5 ) and ( 6 ) , one finds that eq .
( 10 ) can be re - written as @xmath120 for any integer @xmath121 .
averaging this @xmath121 between @xmath122 and @xmath123 and then taking the limit as @xmath124 , it follows that the floquet space detection operator in eq .
( 12 ) can be written equivalently as @xmath125 in matrix form , the detection operator now has the hilbert space - sized matrix @xmath126 in _ all _ diagonal blocks @xmath127 in sec .
iii , we argued that the @xmath76 nearest degeneracies among the floquet states can be condensed into an @xmath75 hilbert space - sized matrix @xmath128 that is infinitely repeated , each copy differing only by a factor of @xmath129 for fourier index @xmath50 .
thanks to the symmetry employed in constructing the alternative form of the floquet space detection operator in eq .
( 40 ) , both it and the initial density matrix are invariant to a change of fourier index .
it therefore follows that the corresponding density ( @xmath130 ) and detection ( @xmath131 ) matrices for each copy of @xmath132 are identical . to build these matrices , we simply insert the matrix elements from @xmath48 and @xmath49 corresponding to those in @xmath132 . as outlined above , in the long rp lifetime regime
, we can approximately evaluate the double sum in the expression for @xmath119 by only including contributions from nearly - degenerate states .
indeed if @xmath133 , we need only be concerned with terms that arise from the near - degeneracies that are captured in each fourier - shifted copy of @xmath128 . to include these terms
, it suffices to diagonalise a single copy of @xmath132 with the fourier index @xmath134 , transform a single copy of @xmath131 and @xmath130 into the eigenbasis ( the eigenvectors of each effective hamiltonian matrix shifted by @xmath135 being the same ) , and calculate the corresponding contributions to @xmath119 .
furthermore , because the expression for @xmath119 only depends on energy _
differences _ , and each copy of @xmath132 is merely shifted in energy by @xmath136 , the contribution to @xmath119 from each nearly - degenerate sub - space with a different fourier index will be the same .
therefore , we have exactly @xmath137 identical contributions to the singlet yield .
this factor cancels with the normalisation constant in the detection operator to leave a double sum over the states in the hilbert space , @xmath138 here @xmath139 is calculated without the factor of @xmath140 in eq .
( 40 ) , and @xmath141 and @xmath142 could be eigenvalues of any copy of @xmath128 , for example that with fourier index @xmath134 .
in order to test the accuracy of this floquet - based approximation , we have carried out calculations on a small model system which could also be simulated exactly using a time - dependent propagation method .
the system is comprised of a single proton on each radical with parallel axial hyperfine interaction tensors @xmath143 mt and @xmath144 mt .
these are based on the largest hyperfine couplings in the fad - tryptophan rp in cryptochrome,@xcite and are therefore representative of the hyperfine interactions that arise in biological systems . given our interest in the magnetic compass sense of migratory birds , which is believed to be associated with the sensitivity of the singlet yield to the direction of an earth - strength ( @xmath145 t ) static magnetic field , we used both eq .
( 41 ) and the exact time - dependent propagation method to compute @xmath146 as a function of the static field direction @xmath147 .
this was done for a variety of radical pair recombination rate constants and both monochromatic and broadband rf radiation in order to assess the accuracy of eq .
( 41 ) in a variety of regimes .
we initially tested the floquet method with a perturbation comprised of a single fourier component @xmath148 and its adjoint @xmath149 here @xmath85 was taken to be the gyromagnetic ratio of a free electron ( @xmath150 mhz/@xmath0 t ) for each radical in the pair .
we arbitrarily selected the frequency @xmath151 mhz , which comes into resonance with a spacing between energy levels of @xmath152 near @xmath153 and @xmath154 .
we also chose an arbitrary initial phase @xmath155 and inclination @xmath156 of the rf magnetic field . between the static magnetic field axis and the rp @xmath157 axis .
here the rf magnetic field strength is @xmath158 nt and the lifetime of the radical pair is @xmath159s . ]
figure 3 compares the singlet yield obtained from eq .
( 41 ) with the exact quantum mechanical singlet yield as a function of @xmath160 for this model problem , with a perturbation strength of @xmath158 nt and a radical pair lifetime of @xmath161s .
the two curves , which each consist of three spikes superimposed on a mildly varying sinusoidal background , are seen to be identical to graphical accuracy . the central spike at @xmath162 in fig .
3 has been discussed in detail in a recent paper.@xcite it arises from a narrowly avoided crossing between the eigenvalues of @xmath163 , which gives rise to a negative lorentzian lineshape in @xmath146 .
the narrowness of the spike has been suggested as a possible explanation for the high precision of the magnetic compass sense of migratory birds.@xcite the additional spikes at @xmath164 and @xmath165 in fig .
3 are generated by the rf field : they occur when the frequency of this field becomes resonant with an energy level spacing of @xmath152 .
these spikes are not purely lorentzian : they have more complicated lineshapes that are sensitive to the initial phase and inclination of the rf magnetic field .
t and ( b ) @xmath166t.,title="fig : " ] t and ( b ) @xmath166t.,title="fig : " ] the results in fig . 3 show that the floquet approximation is essentially exact for this model problem when @xmath167 nt and @xmath168s .
but from the derivation presented in secs .
iii and iv , one would expect it to become less accurate as the strength of the perturbation is increased and the radical pair lifetime is decreased . in order to explore this , we have gone on to investigate what happens when @xmath169 is increased and @xmath170 is decreased by 1 - 2 orders of magnitude . the effect of increasing @xmath169 is shown in fig . 4 .
as the strength of the rf perturbation is increased , the associated resonances at @xmath164 and @xmath165 broaden , while the central spike at @xmath162 remains unchanged .
the broadening of the resonances is captured reasonably well by the floquet approximation up to a magnetic field strength of @xmath171 t .
but by the time @xmath172 t , the approximation has broken down .
since this is already half the strength of the static zeeman field , @xmath173 t , it is not surprising that it is no longer valid to treat it as a perturbation .
the effect of decreasing @xmath170 is shown in fig .
this effect is qualitatively different from that of increasing @xmath169 , in that it decreases the amplitudes of both the rf field - induced resonances at @xmath164 and @xmath165 and the central spike at @xmath162 .
the narrow angular features in the singlet yield are all washed out by the lifetime broadening associated with a shorter - lived rp .
this effect is seen to be captured almost quantitatively by the floquet approximation when @xmath174s , but not for a rp lifetime as short as 1 @xmath0s .
s and ( b ) @xmath175s.,title="fig : " ] s and ( b ) @xmath175s.,title="fig : " ] we have also tested the floquet approximation for a broadband rf perturbation , made up of equally spaced fourier components.@xcite for this , we used a base frequency of @xmath176 khz , including fourier components with @xmath177 between 100 and 2000 .
the band limits were therefore 1.59 and 31.8 mhz . given the much larger number of fourier components in this rf field compared with the monochromatic example considered above , the strength of each component was reduced to @xmath178 nt to give a manageable perturbation .
the phase @xmath97 and inclination @xmath179 of each component were again chosen randomly , and the calculations were performed with a radical pair lifetime of @xmath180 ms .
figure 6 shows the @xmath146 curves for this broadband noise model as calculated with an exact time - dependent propagation method and with the floquet formula in eq .
the singlet yield in the absence of the rf field is also shown for comparison .
although there are now many more resonances than in the monochromatic case , one sees that the agreement between the exact and floquet results is again very good , both in terms of the positions of the rf field - induced resonances and their intensities relative to the rf field - free signal . finally , to test whether the floquet approximation remains reliable for larger radical pairs , we have repeated this broadband noise calculation with an additional proton added to each radical .
the additional proton in the first radical was given the hyperfine tensor @xmath181 and that in the second radical the axial hyperfine tensor @xmath182 , both in mt .
these are again based on the hyperfine interactions of magnetic nuclei in the fad - tryptophan radical pair in cryptochrome.@xcite all the other details of the calculation were kept the same .
the results are shown in fig . 7 , again with the rf field - free signal included for comparison .
now this field - free signal contains three spikes due to avoided crossings between the energy levels of @xmath152 , the original narrow spike at @xmath162 and two broader spikes at @xmath183 and @xmath184 .
the effect of the broadband rf field is seen to be much the same as before , giving rise to a dense forest of narrow resonances on top of the rf field - free background .
the floquet results in fig .
7 are again in good agreement with the exact results in terms of the density of the resonances and their positions , although they are not in quite such good agreement as those in fig . 6 in terms of all of the resonance intensities .
this is slightly concerning for future applications of the floquet method to larger and more realistic radical pairs , but one might at least hope that its predictions would be qualitatively reasonable . since the exact ( time - dependent propagation ) results in fig .
7 took five orders of magnitude more computer time to generate than the floquet results , we believe that there really is no practical alternative to the floquet method for studying long - lived radical pairs with many more nuclear spins in rf magnetic fields .
in this paper , we have shown how a modified formulation of floquet theory combined with degenerate perturbation theory can be used to provide a reasonable approximation to the singlet yield of a radical pair recombination reaction in the presence of a rf magnetic field .
in example calculations of small radical pairs with biologically reasonable hyperfine couplings , the resulting floquet approximation has been shown to be quantitatively accurate for monochromatic rf magnetic fields with @xmath185 t and radical pair lifetimes @xmath186s ( see figs . 3 to 5 ) , and qualitatively reasonable for a broadband magnetic field with @xmath187 fourier components each with @xmath188 nt ( see figs . 6 and 7 ) .
since our implementation of the floquet approximation only involves matrix operations in which the matrices are the size of the hilbert space [ see eq .
( 41 ) ] , its computational cost is only a couple of times larger than that of a standard ( rf field - free ) radical pair singlet yield calculation done by hamiltonian matrix diagonalisation .
a large number of such calculations have been performed in recent years involving radical pairs with as many as twenty or so nuclear spins,@xcite typically by exploiting the separability of the hamiltonian in eq .
any radical pair that has previously been studied in this way could clearly now be exposed to a rf magnetic field and its response studied with the help of the present floquet approximation , including the fad - tryptophan radical pair in cryptochrome that has been suggested as the origin of the magnetic compass sense of migratory birds.@xcite insofar as the disruption of this compass sense by rf radiation is concerned , the present results in figs .
6 and 7 are already rather interesting . if it is true , as suggested in ref .
, that the precision of the compass arises from the narrowness of the spike at @xmath162 ( which is present in the rf field - free signal ) , then one could imagine that the many additional narrow resonances that are generated by the ( broadband ) rf field might well distract the bird from the true north - south axis and lead to disorientation .
however , we should stress that these calculations were performed using vastly over - simplified models of the fad - tryptophan radical pair involving just 2 and 4 nuclear spins , and that the broadband noise used in the calculations had an intensity of @xmath178 nt in each of its @xmath187 fourier components ( giving a root - mean - square fluctuating magnetic field averaged over a cycle of 154 nt ) .
the real fad - tryptophan radical pair contains many more hyperfine - coupled nuclear spins , and experimental investigations of the disorientation of birds by broadband electromagnetic noise have involved rather weaker rf magnetic fields.@xcite both of these factors ( and also the role of electron spin relaxation@xcite ) will have to be taken into account before we can draw any firm conclusions about the effect of rf radiation on the avian compass .
we plan to use the present floquet theory to investigate the effect of weaker rf magnetic fields on more realistic models of the fad - tryptophan radical pair in a future article .
finally , we should point out that we have simply followed earlier work@xcite in assuming that the magnetic component of the rf radiation disrupts the bird s magnetic compass .
this may or may not be the case.@xcite it is also conceivable that the electric component of the radiation plays some role . in any event , we do believe that the theory we have developed here provides a practical way to study the effect of rf magnetic fields on a wide variety of radical pairs that arise in biological and also chemical@xcite systems .
+ we gratefully acknowledge funding from the european research council under the european union s 7th framework programme , fp7/2007 - 2013/erc grant agreement no . 340451 , and from the us air force ( usaf ) office of scientific research under the air force materiel command , usaf award fa9550 - 14 - 1 - 0095 . | we present a new method for calculating the product yield of a radical pair recombination reaction in the presence of a weak time - dependent magnetic field .
this method successfully circumvents the computational difficulties presented by a direct solution of the liouville - von neumann equation for a long - lived radical pair containing many hyperfine - coupled nuclear spins . using a modified formulation of floquet theory , treating the time - dependent magnetic field as a perturbation , and exploiting the slow radical pair recombination
, we show that one can obtain a good approximation to the product yield by considering only nearly - degenerate sub - spaces of the floquet space . within a significant parameter range ,
the resulting method is found to give product yields in good agreement with exact quantum mechanical results for a variety of simple model radical pairs .
moreover it is considerably more efficient than the exact calculation , and it can be applied to radical pairs containing significantly more nuclear spins .
this promises to open the door to realistic theoretical investigations of the effect of radiofrequency electromagnetic radiation on the photochemically induced radical pair recombination reactions in the avian retina which are believed to be responsible for the magnetic compass sense of migratory birds . |
after a meal or a drink , fidgety people sometimes play with the available props , such as empty glasses , bottles , and soda cans .
these containers are all nearly axi - symmetric objects with round bottoms .
a natural form of play is to roll these containers on their circular bottoms . when a slightly tipped container ( fig .
[ fig : ovaltinetippedphoto ] ) is let go , sometimes one sees it fall upright , make a slight banging sound , and then tip up again at another angle , then fall back upright again , and so on for a few rocking oscillations . because these oscillations usually damp quickly its easy to miss the details of the motions . to aid the eye , instead of just letting go of the slightly tilted container , you can flick its top forward with the fingers .
this provides an initial righting angular velocity along the axis about which the container was initially tipped ( too large an angular velocity will cause the container to lift off the table as it pivots over ) .
the tipping container , as before , falls to a vertical configuration , at which time its bottom circular face bangs on the table .
after this bang , the container tips up onto the other side and maybe falls over . when this experiment is performed with a container that is not too tall and too thin
, you can see that the container does not fall exactly onto the diametrically opposite side of the bottom rim .
that is , point a on the bottom of the container that initially contacted the table and the new contact point b , that the container rocks up onto , are not exactly 180 degrees apart .
this experiment ( video available online ) is shown schematically in fig .
[ fig : fallsequence ] . fig .
[ fig : scatterplotovaltine ] shows a histogram of one particular container s orientations after we repeatedly flicked it .
the distribution is strongly bimodal with no symmetric falls over many repeated trials . and [ fig : scatterplotovaltine ] . ]
note the apparent symmetry breaking with @xmath1 .
is this deviation from symmetric rocking due to imperfect hand release ?
here we show that the breaking of apparent symmetry is consistent with the simplest deterministic theories , namely smooth rigid body dynamics .
we derive formulas for the `` angle of turn '' , for both the perfect - rolling and for the frictionless - sliding cases given infinitesimal symmetry breaking in the initial conditions .
the results here generalize some of the nearly - falling - flat results of cushman and duistermaat @xcite ; they considered the special case of the pure - rolling of flat disks . .
see fig .
[ fig : scatterplotovaltine ] ]
if the table was perfectly planar and horizontal , the container s bottom was perfectly circular , and the container was perfectly axisymmetric , then an initial condition with a purely righting angular velocity would result in a collision in which , just as the container becomes vertical , all points on the container s bottom slap the table - top simultaneously .
the consequence of such a rigid - container collision is not computable since algebraic rigid - body collision laws are not well - defined for simultaneous multi - point collisions ; for example , the order of the impulse locations is then ill - defined @xcite . basing the rocking outcome on such a perfect flat collision
would be basing the outcome on details of the deformation , that is , leaving the world of rigid - object mechanics .
geometric perturbations to the circular rim can similarly lead to two - point collisions , where both points of contact are now on the rim .
the second point of contact could be diametrically opposite to the first point , or anywhere else on the rim .
we could then compute the consequences of , say , a plastic collision at the new point of contact .
however , the consequences of the collision would depend on the location of the geometric imperfection . for a given collision point ,
such a theory could predict the energy dissipation at the collision .
but such a theory can not be useful in predicting a systematic breaking of symmetry as seen in fig .
[ fig : scatterplotovaltine ] . for
any given imperfection , the motion is deterministic and depends on the location of the impection , and there is no reason to expect that the location of the imperfection would have a distribution similar to that in fig . [
fig : scatterplotovaltine ] .
degrees is strongly bimodal .
whether the fall is to the left or to the right depends sensitively on initial conditions .
the leftward falls have a mean of about @xmath2 degrees and the rightward falls have a mean of about @xmath3 degrees . for this cylinder theory
predicts @xmath4 degrees for frictionless sliding and @xmath5 degrees for rolling with no slip . ]
assuming a perfectly flat rigid bottom , the circle - slapping - ground simultaneous collision is essentially impossible . after all , the container is being launched by imperfect human hands that can not provide any exact initial conditions
. accounting for various symmetries in the problem , the set of all motions of a container rolling without slip is three dimensional ; the space of solutions could be parameterized by , say , the minimum tip angle @xmath6 , the yaw rate @xmath7 at that position , and the rolling rate @xmath8 at that position .
the set of solutions that leads to a face - down collision can be characterized with only two parameters , a set with co - dimension one @xcite .
so , small generic perturbations of a `` collisional '' initial condition result , with probability 1 , in a non - collisional motion described by the smooth dynamics of the container rolling or sliding on its circular bottom rim .
the rest of this paper is about the near - collisional motions .
we will see that these near - collisional motions involve rapid rolling or sliding of the container on its bottom rim , so that the contact point appears to have switched by a finite angle that is greater than 180 degrees . for this analysis
, we assume a geometrically perfect container ( axisymmetric ) and table ( flat ) .
consider a container with mass @xmath9 , bottom radius @xmath10 , and the center of mass at a height @xmath11 from the bottom .
the moment of inertia is @xmath12 about it s symmetry axis and is @xmath13 about any axis passing through the center of mass and perpendicular to the symmetry axis . for the disk of @xcite , @xmath14 and @xmath15 . in our case , @xmath16 and @xmath17 .
the center of mass of the container is at @xmath18 in an inertial frame @xmath19-@xmath20-@xmath21 at rest with respect to the table .
the reference orientation of the container is vertical as shown in fig .
[ fig : eulerangles]a . any other orientation of the container can be obtained from the reference orientation by a sequence of three rotations , defining corresponding euler angles as illustrated in fig .
[ fig : eulerangles ] . the first rotation ( ` yaw ' or ` steer ' ) is about @xmath21 by an angle @xmath22 .
this rotation also transforms the inertial coordinate axes to @xmath23-@xmath24-@xmath25 . the second rotation ( ` pitch ' or ` tilt ' ) by an angle @xmath26 about the @xmath24 axis results in the @xmath27-@xmath28-@xmath29 frame and determines the orientation of the container up to a rotation @xmath30 about the body - fixed symmetry axis ( @xmath29 axis ) .
the angular velocity of the container in the rotating @xmath27-@xmath28-@xmath29 frame is therefore entirely along the symmetry axis @xmath29 and this relative angular velocity magnitude is denoted by @xmath8 .
we will mostly consider two simple extremes for the frictional interaction between the table and the container s bottom , namely , sliding without friction and rolling without slip . for pure rolling
, the euler angles and their first and second time derivatives determine the center - of - mass position ( relative to the contact point ) , the center - of - mass velocity and acceleration , as well as the container s angular velocity and angular acceleration .
the three second order odes that determine the evolution of the orientation @xmath22 , @xmath31 , and @xmath30 follow from angular momentum balance about the contact point : @xmath32 where @xmath33 these equations are derived in appendix [ app : eqmotion ] . because the center - of - mass velocity is determined by the euler angles and their rates , once these are known the center of mass position can be found by integration . for the frictionless sliding of the cylinder on its circular bottom
, the equations of motion are also given by eq .
[ eq : generalcyl ] but with @xmath34 when the table is frictionless , the horizontal velocity of the center of mass is a constant , so that the horizontal position is independent of the orientation .
the vertical position of the center of mass is simply given by @xmath35 .
without loss of generality the center - of - mass can be taken as on a fixed vertical line .
our initial discovery of the phenomenon described here was found in numerical simulations of the equations above .
we used an adaptive time - step , stiff integrator because the solutions of interest have vastly changing time - scales .
the singularity of the euler - angle description in the near - vertical configuration also contributes to the stiffness of the equations . and
used for the experiments in fig .
[ fig : scatterplotovaltine ] : @xmath36 in consistent si units .
initial conditions @xmath37 .
all the angles are in radians .
the step - change in @xmath22 is the angle of turn @xmath38 , and is a little more than @xmath0 as noted in the caption for fig .
[ fig : scatterplotovaltine ] . ] for simplicity of presentation , we first describe in detail the results of integration when the container does not slip with respect to the table .
first , we integrate the equations with initial conditions that lead exactly to a face - down collision @xmath39 , @xmath40 , and say , @xmath41 . as would be expected , the container lands in a manner that all of its bottom face simultaneously reaches the horizontal table - top .
up to this time the motion of the container is identical to that of an inverted pendulum hinged at the contact point a. the integration becomes physically meaningless at that collision point . in the vicinity of these collisional motions , the solutions of the equations of motion here
do not have smooth dependence of solutions on initial conditions . to obtain a near - collisional motion , we set @xmath42 , @xmath43 , @xmath40 , and @xmath44 , where @xmath45 is a small quantity
. the results of the integration are shown in fig .
[ fig : purerollnumresults ] .
the plot of @xmath46 suggests a motion in which the container s bottom face periodically comes close to touching the table ( @xmath47 ) , but then gets `` repelled '' by the floor as if by an elastic collision , so that the container rocks down and up periodically , ad infinitum , without losing any energy , as expected from this dissipation - free system .
we notice that when @xmath48 , @xmath49 changes almost discontinuously . also , when @xmath31 is close to zero , both @xmath7 and @xmath8 blow up to very large values , resulting in almost discontinuous changes in the corresponding angles @xmath22 and @xmath30 .
note that we are simply simulating the apparently smooth differential equations , and not applying an algebraic transition rule for a collision .
as @xmath45 goes to zero , the angle rates @xmath7 and @xmath8 grow without bound , but the magnitude of the angular velocity vector is always bounded , as it must be since energy remains a constant throughout the motion . in particular , while the angle rates @xmath8 and @xmath7 are large , they are very close to being equal and opposite ( @xmath50 , fig . [
fig : purerollnumresults]i ) .
let us examine the consequences of a rapid finite change in @xmath22 when @xmath48 .
the position of the contact point p@xmath51 on the ground relative to the center of mass @xmath18 is given by the following equations : @xmath52 when @xmath48 and @xmath22 is not particularly close to any multiple of @xmath53 , the contact point position is given by @xmath54 given that @xmath55 do not change much during the brief near - collisional phase ( because center of mass velocity is finite ) , we can see from eq .
[ eq : contactptcircle ] that a rapid continuous change in @xmath22 corresponds to a rapid continuous change in the contact point in a circle with the center @xmath56 .
thus the `` angle of turn '' aob defined earlier is simply the change in @xmath22 . at the singular limit of a near - collisional motion arbitrarily close to a collisional motion
, this continuous but steep change in @xmath22 approaches a step change this is the `` limiting angle of turn '' and we denote this by @xmath57 .
we consider the near - collisional motions of the no - slip container that can be characterized as being the pasting - together of two qualitatively distinct motions of vastly different time - scales : 1 .
inverted pendulum - like motion about an essentially fixed contact point when the tilt angle @xmath31 is large .
rapid rolling of the container which accomplishes in infinitesimal time , a finite change in the contact point , and a sign - change in the tilt rate @xmath49 .
note that the near - collisional motion for a container rolling without slip requires very high friction forces ( for non - zero @xmath11 , see fig .
[ fig : purerollnumresults]g ) .
however , plotting the ratio of the required friction forces with the normal reaction , we find that only a finite coefficient of friction is required for preventing slip even in the collisional limit ( fig .
[ fig : purerollnumresults]h ) . the other extreme of exactly zero friction is similar . here
, the horizontal component of the center of mass velocity may be taken to be zero .
the near collisional motions for a container sliding without friction are again characterized as consisting of two qualitatively different phases : 1 . a tipping phase when @xmath31 is not too small , involving the container moving nearly in a vertical plane , the center of mass moving only vertically , and the contact point slipping without friction .
a rapid sliding phase in which the sign of @xmath49 is reversed almost discontinuously , and the contact point moves by a finite angle in infinitesimal time .
the angle of turn for the frictionless case and the no - slip case are different in general ( when @xmath58 ) . in the next two sections , we derive the formulas for the angle of turn by taking into account the two - phase structure of the near - collisional motion .
the rigid body dynamics of disks , containers , and similar objects with special symmetries , have been discussed at length by a number of authors , including distinguished mechanicians such as chaplygin , appell , and korteweg .
their works include complete analytical characterizations of the solutions to the relevant equations of motion , typically involving non - elementary functions such as the hyper - geometric .
reasonable reviews of such literature can be found , for instance , in @xcite and @xcite .
we will not use these somewhat cumbersome general solutions but will analyze only the special near - collisional motion of interest to us .
the calculation below may be called , variously , a boundary layer calculation , a matched asymptotics calculation ( but we are not interested in an explicit matching ) or a singular perturbation calculation . essentially , we take advantage of the presence of two dynamical regimes with vastly different time - scales , each regime simple to analyze by itself . the overall motion can be obtained approximately by suitably pasting together the small-@xmath31 and the large-@xmath31 solutions .
but the angle of turn is entirely determined by the small-@xmath31 regime , as will be seen below .
first , consider eq .
[ eq : generalcyl ] with @xmath59 in the limit of small @xmath31 , so that we can use @xmath60 and @xmath61 , and generally neglect terms of @xmath62 .
we obtain @xmath63 now , considering eq .
[ eq : generalcyl ] with @xmath64 in the limit of small @xmath31 , we obtain @xmath65 eqs .
[ eq : eq1atsmallphi ] and [ eq : eq3atsmallphi ] are linear and homogeneous in @xmath7 , @xmath8 , and their time derivatives .
so , positing a linear relation between @xmath7 and @xmath8 , we find that the following two equations are equivalent to eqs .
[ eq : eq1atsmallphi ] and [ eq : eq3atsmallphi ] : @xmath66 and @xmath67 note that eq .
[ eq : psieqtheta ] agrees with the results of the numerical simulation , as in fig .
[ fig : purerollnumresults]i .
even though this equation was derived for small @xmath31 , this equation is approximately true at large @xmath31 as well if the initial conditions of the motion at large @xmath31 have @xmath68 and @xmath43 , as is the case for the near - collisional motions we consider . integrating eq .
[ eq : noslip1 ] , we obtain @xmath69 thus , when @xmath70 , both @xmath7 and @xmath8 ( @xmath71 ) become very large .
now consider eq .
[ eq : generalcyl ] with @xmath72 and with @xmath70 .
@xmath73 using eq .
[ eq : psieqtheta ] in eq .
[ eq : eq2atsmallphi ] and ignoring higher order terms in @xmath31 , we obtain @xmath74 using eq .
[ eq : psidotandphi ] for @xmath7 in the above equation and neglecting the @xmath75 term ( as it is a constant and therefore , much smaller than 1/@xmath76 for small @xmath31 ) , we get @xmath77 where @xmath78 the general solution for the differential equation in @xmath31 , eq .
[ eq : rollphidotdoteqn ] , can be written as @xmath79 where @xmath80 is the lowest @xmath31 attained by the container before rising back up again , and @xmath81 is when this minimum @xmath31 is attained . substituting this equation in eq .
[ eq : psidotandphi ] , we obtain a simple equation for the evolution of @xmath22 when near the surface : @xmath82 we can now compute the change in @xmath22 during a small time interval @xmath83 centered at @xmath84 .
@xmath85 from conservation of energy , we can show that @xmath86 , and therefore @xmath87 , must be @xmath88 ( see appendix [ app : b2scaling ] ) .
hence , @xmath89 as @xmath90 . keeping @xmath83 a small constant as we let @xmath90 , thus approaching the collisional limit , gives us the following expression for @xmath38 .
@xmath91 this is the limiting angle of turn when the container rocks by rolling without slip .
earlier numerical integrations agree quite well with this formula near the collisional limit , as they should .
we now briefly outline the procedure for deriving the angle of turn for the frictionless case .
the procedure closely parallels that described for the no - slip case , except for small differences below .
the frictionless equations eq .
[ eq : generalcyl ] and eq .
[ eq : nofric_eqs ] corresponding to @xmath64 simplifies to @xmath92 here , @xmath93 because the initial conditions satisfy @xmath43 and @xmath68 by assumption , as before . using eq .
[ eq : slipangmom ] in the frictionless equation eq .
[ eq : generalcyl ] and eq .
[ eq : nofric_eqs ] corresponding to @xmath59 , we have , after some simplifications : @xmath94 where @xmath86 is a constant of integration , whose order is estimated in appendix [ app : b2scaling ] . substituting this into the @xmath95 frictionless equation and simplifying by neglecting all higher order @xmath31 terms
, we eventually obtain @xmath96 using arguments identical to the pure - rolling case , this results in the following expression for the angle of turn , @xmath97 for the frictionless limit .
in this section , we derive the same angle of turn formulas ( eq . [ eq : noslipangleofturn ] and [ eq : nofricangleofturn ] ) without referring back to the complicated full dynamical equations .
rather , using heuristic reasoning , we directly derive equations of motion that apply at the small angle limit ( @xmath98 ) .
we represent the lean of the cylinder by the vector @xmath99 with magnitude equal to @xmath31 and direction along the @xmath24 axis : @xmath100 .
these quantities @xmath101 and @xmath22 are related to each other exactly like @xmath102 and @xmath30 , respectively , in traditional polar coordinates . neglecting any angular velocity component along @xmath103 ,
the angular velocity vector for the cylinder is given by @xmath104 .
the rate of change of angular momentum @xmath105 about the center of mass g is given by : @xmath106 the angular momentum balance equation is then given by @xmath107 where @xmath108 is the moment of all the external forces about g. @xmath108 depends on whether or not there is friction ; so we treat the two cases in turn .
the vertical position of the center of mass is @xmath109 .
so @xmath110 .
the vertical ground reaction is thus @xmath111 , neglecting gravity @xmath112 in comparison .
the moment @xmath108 of this vertical force about the center of mass g is equal to @xmath113 in the direction @xmath24 . substituting this along with eq .
[ eq : heuristic_angmom ] in the angular momentum balance equation eq .
[ eq : angmombalance ] , we have @xmath114 this vector equation is identical to eqs .
[ eq : nofric1 ] and [ eq : nofric2 ] and therefore , lead to the same angle of turn ( eq . [ eq : nofricangleofturn ] ) .
the position of the center of mass g with respect to the point p@xmath115 on the cylinder in contact with ground is given by @xmath116 .
the velocity of the center of mass g is given by @xmath117 .
using the no - slip constraint @xmath118 , we obtain after some simplifications , the acceleration of the center of mass to first order to be @xmath119 the ground reaction force , which now includes a horizontal friction force as well , is simply @xmath120 , neglecting gravity .
the moment of this ground reaction force about g is given by @xmath121 equating @xmath108 above to @xmath122 from eq .
[ eq : heuristic_angmom ] gives eqs .
[ eq : noslip1 ] and [ eq : ddphianddpsi ] from the previous version of the derivation , therefore resulting in the same formula for the angle of turn ( eq . [ eq : noslipangleofturn ] ) .
firsly and most significantly , note that the limiting angle of turn does not depend on the initial conditions such as the initial tilt and tip velocity .
this means that we do not have to control these accurately in an experiment .
this also agrees with the relatively small variance in the histogram of fig .
[ fig : scatterplotovaltine ] , in which we did not control the initial conditions . the histogram fig .
[ fig : scatterplotovaltine ] was obtained using the cylindrical container shown in fig .
[ fig : ovaltinetippedphoto ] with @xmath123 cm , @xmath124 cm , @xmath125 m@xmath126 . using these numbers in the angle of turn formulas
gives an angle of turn of about 220 degrees for frictionless sliding and an angle of turn of @xmath127 degrees for rolling without slip .
these angles of turn would manifest as a deviation of either @xmath128 degrees or @xmath129 degrees from falling over to exactly the diametrically opposite side . in the toppling experiment of fig .
[ fig : scatterplotovaltine ] , the container orientation was 33 degrees on average from 180 degrees in the leftward falls ( suggesting an angle of turn of 213 degrees ) and was about 37.9 degrees on average from 180 degrees in the rightward falls ( suggesting angle of turn of 217.9 degrees ) .
the standard deviations were respectively 3.9 and 4.5 degrees respectively .
the experimental angle of turn seems better predicted by the asymptotic formula for frictionless sliding in this case .
although , neither the frictionless limit nor the no - slip limit is just right , both limits capture the many qualitative aspects of the motion quite well .
for tall thin cylinders with @xmath130 and @xmath131 , both equations for the angle of turn , eq . [ eq : noslipangleofturn ] and eq .
[ eq : nofricangleofturn ] , tend to @xmath0 radians .
that is , very tall cylinders are predicted to have a smaller symmetry - breaking .
this prediction agrees with the common experience that when we tip a tall - enough cylinder ( something that looks more like a tall thin beer bottle ) in a manner that its bottom surface nearly falls flat , the cylinder essentially rocks up on a contact point almost diametrically opposite to the initial contact point .
thus , this apparently almost - symmetric rocking is accomplished by a rapid asymmetric rolling or sliding of the container over roughly one half of its bottom rim !
this result is basically independent of the container - table frictional properties .
we may define a disk as a container with zero height : @xmath14 , a good approximation is the euler s disk @xcite . if the radius of gyration of a disk is @xmath132 , then @xmath133 and @xmath134 . substituting these into the angle of turn formulas for the no - slip ( eq .
[ eq : noslipangleofturn ] ) or the frictionless ( eq . [ eq : nofricangleofturn ] ) cases , we obtain the same angle of turn : @xmath135 indeed , numerical exploration with eqs .
[ eq : generalcyl]-[eq : nofric_eqs ] , suitably modified for frictional slip and specialized to disks , shows no dependence of the angle of turn @xmath136 on the form of the friction law or the magnitude of the friction .
this lack of dependence on friction could be anticipated from our small angle calculations for pure rolling cylinders , specialized to disks .
in particular , we find that the acceleration of the center of mass for a pure rolling cylinder ( eq . [ eq : rgdotdot ] ) with @xmath14 is vertical in the small - angle limit , enabling the no - slip condition to be satisfied even without friction .
thus , the rolling solution is obtained with or without friction .
for the special case of pure - rolling disks , eq .
[ eq : diskangleofturn ] was found in @xcite . *
a homogeneous disk * has @xmath137 .
the corresponding angle of turn is equal to @xmath138 , which is about 41 degrees more than a full rotation of the contact point .
this prediction is easily confirmed in casual experimentation with metal caps of large - mouthed bottles or jars on sturdy tables for such caps , we observe that the new contact point is invariably quite close to the old contact point . *
a ring * such as the rim of a bicycle wheel has @xmath139 .
the corresponding angle of turn is equal to @xmath140 , which is about 48 degrees less than a full rotation of the contact point .
thus the apparent near - collisional behavior of a homogeneous disk and a ring will be superficially similar , even though the actual angles of turn differ by about @xmath141 degrees . the angle of turn can be controlled by adjusting @xmath142 .
for instance , between a ring and a disk is an object that appears to bounce straight back up the way it falls . and
it is possible to increase the theoretical angle of turn without bound by choosing @xmath143 , a disk in which almost all the mass is concentrated at the center .
here we analyzed what happens when a cylinder or a disk rocks to an almost flat collision on its bottom surface .
we found that the smallest deviation from a perfect face - down collision of a container s bottom results in a rapid rolling and/or sliding motion in which the contact point moves through a finite angle in infinitesimal time .
calculations of this finite angle explain certain apparent symmetry breaking in experiments involving rocking or toppled containers . in this system ,
the consequences of such a degenerate ` collision ' are a discontinuous dependence on initial conditions ( rolling left or rolling right depend on the smallest deviations in the initial conditions ) . such discontinuous dependence on initial conditions or geometry is a generic feature of systems in the neighborhood of simultaneous collisions .
other examples include a pool break or the rolling polygon of @xcite .
thanks to arend schwab for comments on an early manuscript and mont hubbard for editorial comments .
the alternate heuristic derivation of section [ sec : heuristic ] were informed by discussions with anindya chatterjee in the context of the euler s disk .
ms was supported by cornell university in the early stages of this work ( 2000 ) .
the work was also supported partly by nsf robotics ( cise-0413139 ) and fibr ( ef-0425878 ) grants .
the moment of inertia tensor of the cylinder about its center of mass is , in dyadic form , @xmath144 the angular velocity @xmath145 of the cylinder is given by @xmath146 the angular momentum about the center of mass is @xmath147 the rate of change of angular momentum @xmath148 is given by the sum of two terms : 1 ) the rate of change relative to the rotating frame @xmath27-@xmath28-@xmath29 , obtained by simply differentiating the components in eq .
[ eq : hg ] , and 2 ) the rate of change @xmath149 due to the rotation of the @xmath27-@xmath28-@xmath29 frame with an angular velocity @xmath150 .
@xmath151 the point s on the cylinder in contact with the ground has zero velocity : @xmath152 , no slip . using @xmath153 ,
the velocity @xmath154 of the center of mass is given by @xmath155 the acceleration of the center of mass @xmath156 is given by adding two terms : 1 ) acceleration relative to the rotating @xmath27-@xmath28-@xmath29 frame , obtained by differentiating the components in eq .
[ eq : vg_noslip ] , and 2 ) an acceleration term @xmath157 due to the rotation of @xmath27-@xmath28-@xmath29 frame .
we obtain @xmath158 the contact force @xmath159 is given by linear momentum balance : @xmath160 then , the moment @xmath161 of all the external forces about the center of mass is @xmath162 in which , upon simplification , we have @xmath163 equating @xmath161 and @xmath164 gives us the angular momentum balance equations of motion for no - slip rolling in eqs .
[ eq : generalcyl],[eq : noslip_eqs ] .
no friction implies that the horizontal velocity of the center of mass g is a constant and can be set to zero by appropriate reference frame choice , without loss of generality .
further , noting that @xmath165 , we have , by differentiating twice : @xmath166 as before , we compute the contact force as @xmath167 and compute the net moment @xmath168 about the center of mass as @xmath169 again , equating @xmath161 with the @xmath164 before gives us the required equations of motion eqs . [ eq : generalcyl],[eq : nofric_eqs ] .
we now establish that @xmath170 , used in obtaining eq .
[ eq : noslipangleofturn ] .
the total mechanical energy @xmath171 of the cylinder , a constant , is given by : @xmath172 in which p.e .
is the potential energy .
we consider this total energy at the time of lowest tip angle @xmath31 , @xmath81 , we have @xmath173 , @xmath174 , @xmath175 , and the potential energy p.e . @xmath176 .
making use of the usual small @xmath31 approximations ( as used in the main text ) and eq .
[ eq : psieqtheta ] , namely @xmath177 , we find that @xmath178 because @xmath171 is a constant , we have @xmath179 or @xmath180 as claimed earlier . the total mechanical energy @xmath171 is now given by @xmath181 again , using small @xmath80 approximations and considering the energy equation at @xmath84 , we obtain @xmath180 , using arguments identical to the no - slip rolling case . | a beer bottle or soda can on a table , when slightly tipped and released , falls to an upright position and then rocks up to a somewhat opposite tilt .
superficially this rocking motion involves a collision when the flat circular base of the container slaps the table before rocking up to the opposite tilt . a keen eye notices that the after - slap rising tilt is not generally just diametrically opposite the initial tilt but is veered to one side or the other .
cushman and duistermaat ( 2006 ) recently noticed such veering when a flat disk with rolling boundary conditions is dropped nearly flat . here , we generalize these rolling disk results to arbitrary axi - symmetric bodies and to frictionless sliding .
more specifically , we study motions that almost but do not quite involve a face - down collision of the round container s bottom with the table - top .
these motions involve a sudden rapid motion of the contact point around the circular base .
surprisingly , like for the rolling disk , the net angle of motion of this contact point is nearly independent of initial conditions .
this angle of turn depends simply on the geometry and mass distribution but not on the moment of inertia about the symmetry axis .
we derive simple asymptotic formulas for this `` angle of turn '' of the contact point and check the result with numerics and with simple experiments . for tall containers
( height much bigger than radius ) the angle of turn is just over @xmath0 and the sudden rolling motion superficially appears as a nearly symmetric collision leading to leaning on an almost diametrically opposite point on the bottom rim . |
our experiments are performed using 87-rubidium atoms in magneto - optical traps ( mots ) that are excited to @xmath0 rydberg states using a two - step excitation process with two lasers at 420 nm and 1013 nm , respectively .
the first step laser is detuned sufficiently from the intermediate 6p state ( by about 2 ghz ) to ensure that the intermediate step is not populated on the timescale of the experiment ( for our scheme , this could lead to a two - step ionization ) .
the laser at 1020 nm is focused to a waist of 80 microns , and the waist of the 420 nm laser can be varied between 6 and 40 microns , allowing us to change the overlap between the atomic clouds of size @xmath31 and the excitation lasers , and hence the effective excitation volume , which can be varied between @xmath32 and @xmath33 .
we note here that for the smallest waist of the 420 nm laser , the resulting excitation geometry is quasi one - dimensional since the waist of the laser is then smaller than the dipole blockade radius .
the atomic clouds have peak densities around @xmath34 . after an excitation pulse of duration between 0.5 and 950 microseconds , during which the mot beams are switched off ,
the rydberg atoms are field ionized by applying a voltage to electrodes located outside the vacuum chamber , and the resulting ions are accelerated towards a channeltron charge multiplier , whose output is recorded on an oscilloscope .
the overall detection efficiency @xmath35 for the ions was determined to be around 40 percent .
we extract information on the rydberg excitations present after a single excitation sequence by recording the channeltron signal due to the arrival of single ions on a digital oscilloscope and then using a peak - finding routine in order to determine the number of ions .
this procedure is repeated up to @xmath36 times , and from the resulting counting distributions the mean @xmath37 as well as the @xmath38-th order central moments @xmath39 ( up to @xmath40 ) are calculated . from those moments , we also calculate the normalized quantities characterizing the distributions : mandel @xmath27-factor @xmath41 , skewness @xmath42 , binder cumulant @xmath43 and bimodality coefficient @xmath44 .
we compare our experimental results to three simple models in order to highlight the main features of the dependence of the various quantities on @xmath37 : a perfectly monomodal and a poissonian monomodal model with mean @xmath45 and central moments @xmath46 for the perfectly monomodal model and @xmath47 , @xmath48 , @xmath49 , and @xmath50 for the poissonian model , as well as a bimodal model with two modes at @xmath51 and @xmath52 and probabilities @xmath53 and @xmath54 , respectively .
the mean and the central moments for the bimodal model are @xmath55 and @xmath56 .
+ the finite detection efficiency @xmath35 in the experiment , which transforms the actual counting distribution @xmath57 into an observed counting distribution @xmath58 , is taken into account by using the analytical result @xmath59 for the @xmath38-th factorial moments @xcite and the relationship between the factorial and central moments given by @xcite @xmath60\mu^i \mu_j^f,\ ] ] where @xmath61 is the stirling number of the second kind . the observed mean and the second , third and fourth central moments
are then related to the actual moments by @xmath62\mu_2+(3\eta^4 - 6\eta^3 + 3\eta^2)\mu^2 + \nonumber \\&&+(-6\eta^4 + 12\eta^3 - 7\eta^2+\eta)\mu\end{aligned}\ ] ] from which the observed values of the normalized quantities can be calculated ( for the mandel @xmath27-factor , the particularly simple result @xmath63 holds ) .
alternatively , the above procedure can be used to calculate the actual moments and normalized quantities from the observed ones . + the numerical simulations shown in fig .
2 are based on an extension of the dicke model applied to rydberg excitations @xcite which is explained in detail in the supplementary information .
briefly , the original dicke model , based on the laser excitation conserving the cooperative character of the collective atomic wavefunctions , was modified by including the van - der - waals interactions between the collective dicke states , taking into account the @xmath64 contribution at short distances for @xmath65 states .
the collective dicke states contain the full statistical information about the collective rydberg excitation and therefore allow the calculation of all the moments of the excitation statistics .
the improved dicke model used here takes into account the time - dependent coupling between the dicke states , which becomes increasingly important as the number of rydberg excitations increases . for large excitation numbers the non - symmetric dicke states experience energy shifts and decoherence due to the van - der - waals interactions .
these , in turn , lead to the asymmetric lineshapes and the characteristic histograms , featuring bimodal distributions for positive detuning , as shown in fig .
the main elements of the dicke model describing the van - der - waals blockaded rydberg excitation of @xmath66 atoms have been exposed in @xcite . here
we will concentrate on the details of the revised model that are relevant for the features observed in the experiment .
the dicke model is based on the laser excitation conserving the cooperative character of the collective atomic wavefunctions .
the description of the system through the collective dicke states contains the full statistical information , not only the average quantities .
all the moments of the excitation statistics can be calculated , in particular the measured full counting distributions of the rydberg excitations . in order to describe the experimental results presented in this paper the dicke model treatment
was improved . as explained in @xcite , the introduction of an orthogonal basis for diagonalizing
the van der waals basis coupling matrix in the restriction of the space of the non - fully symmetrical dicke states allow us to treat the evolution of the system with an single integro - differential equation for each symetrical dicke state with @xmath67 rydberg excitations @xmath68 where @xmath69 is the amplitude of the dicke state in the collective wave function , @xmath1 is the laser detuning , @xmath70 is the rabi frequency for the laser excitation and @xmath71 is the mean value of the van der waals interaction for the atoms in the symmetrical @xmath67 dicke state .
the term @xmath72 in the equation corresponds to the mean field seen by the atoms .
the term @xmath73 is a kernel function defined as@xmath74 \exp\left ( -iw_{qq}\tau\right ) \right]\ ] ] where @xmath75 corresponds to the ensemble of the @xmath76 van der waals states with @xmath67 rydberg excitation and an energy @xmath77 .
the evaluation of the kernel function is a crucial and non - trivial step in solving these equations @xcite . nevertheless , in ref .
@xcite we have shown that , to first approximation , the kernel function term firstly compensates the mean - field term and secondly acts as a coupling between the fully symmetrical @xmath67 dicke state and one particular virtual ad - hoc state with the same number of rydberg excitations , simulating the coupling with the ensemble of the non - symmetrical dicke states .
this previous approach was limited to short interaction times @xmath78 and small numbers of rydberg excitations @xmath79 .
the striking features of the collective dynamics appear at longer interaction times , when the organisation of the rydberg excitations within the ultracold cloud has reached a stationary state .
a more complete treatment needs to take into account the time - dependent coupling between the different states , which becomes more and more important when @xmath79 increases .
the kernel function depends on the distribution of the atoms inside the experimental sample .
analytical solutions can be found for many configurations , but the exact solution of the equations is quite involved .
it is easier , and more interesting for a physical understanding of the problem , to consider the fourier transform @xmath80 of the above kernel , which for a homogeneous atomic density is given to a good approximation by@xmath81 with the relation@xmath82 where @xmath83 is related to the volume of the atomic sample , @xmath84 corresponds to an effective minimum volume given by the mean distance between two atoms and @xmath85 is the van der waals coupling .
this coupling can be estimated by taking into account the @xmath86 contribution at short distances for the @xmath65 states @xmath87 where we introduce an effective volume @xmath88 as@xmath89 @xmath90 corresponds to the minimum energy of @xmath67 rydberg atoms contained in the volume @xmath91 .
it needs to be estimated for each @xmath67 .
the general shape of the fourier transform function is shown in fig . 5 for @xmath92 and @xmath93 .
it is compared with a function calculated by considering an atomic density given by the expression@xmath94 \exp\left [ -\frac{v_{x}^{2}}{v^{2}}-\frac{v_{0}^{2}}{v_{x}^{2}}\right ] , \label{atdensity}\ ] ] where @xmath95 , with @xmath96 the distance between two atoms .
the choice of this density function , which is not far from a homogeneous atomic density in a volume @xmath91 , allows us to pursue the analytical approach by using the expression@xmath97 \exp\left [ -\frac{\omega_{vj\max}}{\omega}-\frac { \omega}{\omega_{vj\min}}\right ] \frac{1}{\omega^{3/2}}\ ] ] both functions exhibit three distinct parts corresponding to very different frequency ranges . for the high frequencies ( @xmath98 ) ,
the shape of the function is broad , corresponding to a markovian short memory process , which in first approximation compensates the mean - field term . for the low frequencies ( @xmath98 ) ,
the function is much narrower , which characterizes a long memory process equivalent to the coupling to a bound state . for the frequencies @xmath99 , which corresponds to the van der waals coupling value , the function is equal to zero . for large @xmath67 ,
this behavior is slightly modified .
the mean field term is partly restored , while the coupling to the the virtual state is decreased in amplitude .
a decay term appears in the equations corresponding to a decoherence process for the non symmetrical dicke states .
the state of the system corresponding to @xmath67 rydberg excitations is described by two differential equations@xmath100 and@xmath101 @xmath102 and @xmath103 are functions such as@xmath104 and@xmath105 @xmath106 is given by@xmath107 with @xmath108 . at this point the dynamics of the system
is no longer hamiltonian . in effect
, we have introduced a decoherence process in the ensemble of the non - symmetrical states , which leads us to consider a population equation to conserve the ensemble of the atoms@xmath109 \left\vert b_{j}\left ( t\right ) \right\vert ^{2}.\ ] ] the non coherent population with @xmath67 rydberg excitations corresponds to a total van - der - waals coupling energy of @xmath110 , which characterizes the global correlation of the system .
the rydberg excitation of the incoherent and non - symmetrical dicke states is treated perturbatively through@xmath111 \right ] \left\vert b_{j}\left ( t\right ) \right\vert ^{2}-g(j)\rho _ { j}+g\left ( j-1\right ) \rho_{j-1}.\ ] ]
these equations depend on very few adjustable parameters , essentially the frequencies @xmath112 and @xmath113 , for which we can find reasonable estimates based on physical arguments . in the case of the @xmath65 states of rubidium which do not present any degeneracy ,
the experimental parameters can be directly used .
compared to the previous treatment in @xcite the dynamics of the evolution is taken into account more carefully and two important new features have been introduced .
the first one is the decoherence in the evolution , and the second one is the appearance of the correlation corresponding to the energy of the virtual state .
the rydberg excitation of the coherent non symmetrical state can be taken into account perturbatively , which does not introduce a fundamental change in the results of the calculations .
we do not expect our model to perfectly reproduce the experimental results , as several details of the experiment ( inhomogeneous density of the cloud , other rydberg levels , decay from the excited state , finite velocity of the atoms ) are not taken into account . also , in our numerical simulations we found that the dynamics of the system depends on several parameters , a slight change of which leads to significatively different results .
furthermore , for long excitation times the results depend sensitively on how we take into account the finite coherence time of the laser excitation .
nevertheless , fig .
2 in the main text shows that for a reasonable choice of parameters in the simulation , which are not too far from the parameters of the experiment , the results of the simulations agree quite well with the experimental data .
in particular , agreement at the level of the full counting statistics ( histograms in fig .
2 ( a - c ) ) is satisfactory . for positive detuning
the histogram exhibits a bimodal structure , while a single peak is essentially observed at resonance and very few atoms are excited for negative detuning .
further information is obtained by analyzing the dynamics of the system . on resonance the atoms
are essentially described by the incoherent anti - symmetrical states , whereas for negative detuning only fully symmetrical dicke states are populated . on resonance
the evolution quickly becomes incoherent , and the system is described by a statistical mixture of states . for negative detuning , by contrast , the system is described by a superposition of a small number of dicke states with little excitation .
interestingly , on resonance the total energy of the different non - coherent antisymmetrical states is low , meaning the system exhibits strong correlations .
the histogram for positive detuning contains both components , resulting in @xmath114 statistical mixture of states and @xmath115 coherent superposition of dicke states .
we also observe that for long excitation times the coherent antisymmetrical states are marginally populated for negative or positive detuning , while they remain present for much longer in the case of resonant excitation .
this work was supported by prin and the eu marie curie itn coherence .
the authors thank r. fazio , a. tomadin , m. dellorso and r. mannella for discussions .
99 verstraete , f. , wolf , m. m. & cirac , j.i . quantum computation , quantum state engineering , and quantum phase transitions driven by dissipation._nat .
phys . _ * 5 * , 633 ( 2009 ) .
weimer , h. , ller , m. m. , lesanovsky , i. , zoller , p. & bchler , h.p . a rydberg quantum simulator . _ nat . phys . _ * 6 * , 382 ( 2010 ) .
ates c. , olmos b. , garrahan j. p. & lesanovsky i. , dynamical phases and intermittency of the dissipative quantum ising model .
a _ * 85 * , 043620 ( 2012 ) .
lee , t. e. , hffner , h. & cross m. c. antiferromagnetic phase transition in a nonequilibrium lattice of rydberg atoms .
a _ * 84 * , 031402(r ) ( 2011 ) .
lee , t. e. , hffner , h. & cross m. c. collective quantum jumps of rydberg atoms .
lett . _ * 108 * , 023602 ( 2012 ) .
hu , a. , lee , t. e. & clark , c. w. spatial correlations of one dimensional driven - dissipative systems of rydberg atoms .
preprint at @xmath116http://arxiv.org / abs/1305.2208@xmath117 ( 2013 ) .
comparat , d. & pillet , p. dipole blockade in a cold rydberg atomic sample . _ j. opt .
b _ * 6 * , a 208 ( 2010 ) .
saffman , m. , walker , t. g. & mlmer , k. quantum information with rydberg atoms , _ rev .
phys . _ * 82 * , 2313 ( 2010 ) .
kessler , e. m. _ et al .
_ dissipative phase transition in a central spin system .
a * 86 * , 012116 ( 2012 ) .
diehl , s. , tomadin , a. , micheli , a. , fazio , r. & p. zoller , dynamical phase transitions and instabilities in open atomic many - body systems .
_ * 105 * , 015702 ( 2010 ) .
schwager , h. , cirac , j. i. & giedke , g. dissipative spin chains : implementation with cold atoms and steady - state properties .
a _ * 87 * , 022110 ( 2013 ) .
carr , c. , ritter , r. , weatherill , k. j. & adams , c. s. cooperative non - equilibrium phase transition in a dilute thermal atomic gas .
preprint at @xmath116http://arxiv.org / abs/1302.6621@xmath117 ( 2013 ) .
schwarzkopf , a. , sapiro , r. e. & raithel , g. imaging spatial correlations of rydberg excitations in cold atom clouds .
lett . _ * 107 * , 103001 ( 2011 ) .
schau , p. _ et al .
_ observation of spatially ordered structures in a two - dimensional rydberg gas .
_ nature _ * 491 * , 87 ( 2012 ) .
viteau , m. _ et al .
_ cooperative excitation and many - body interactions in a cold rydberg gas._phys .
* 109 * , 053002 ( 2012 ) .
braggio , a. , knig , j & fazio , r. full counting statistics in strongly interacting systems : non - markovian effects .
lett . _ * 96 * , 026805 ( 2006 ) .
gogolin a. o. & komnik , a. full counting statistics for the kondo dot in the unitary limit .
lett . _ * 97 * , 016602 ( 2006 ) .
tsai , s .- h .
& salinas , s. r. fourth - order cumulants to characterize the phase transitions of a spin-1 ising model . _
brazilian journal of physics _ * 28 * , 58 ( 1998 ) .
grttner , m. , heeg , k. p. , gasenzer , t. & evers , j. optimal self - assembly of rydberg excitations for quantum gate operations .
preprint at @xmath116http://arxiv.org / abs/1203.2884@xmath117 .
viteau , m. _ et al .
_ rydberg excitations in bose - einstein condensates in quasi - one - dimensional potentials and optical lattices .
lett . _ * 107 * , 060402 ( 2011 ) .
weimer , h. , lw , r. , pfau , t. & bchler , h.p .
, quantum critical behavior in strongly interacting rydberg gases , _ phys .
* 101 * , 250601 ( 2008 ) .
pohl , t. demler , e. & lukin , m. d. dynamical crystallization in the dipole blockade of ultracold atoms .
lett . _ * 104 * , 043002 ( 2010 ) .
schachenmayer , j. , lesanovsky , i. , micheli , a , & and daley , a.j . ,
dynamical crystal creation with polar molecules or rydberg atoms in optical lattices , _ new j. phys . _
* 12 * 103044 ( 2010 ) .
van bijnen , r .
m. w. , smit , s. , van leeuwen , k. a. h. , vredenbregt , e. j. d. & kokkelmans s. j. j. m. f. , adiabatic formation of rydberg crystals with chirped laser pulses .
_ j. phys .
b : at . mol .
phys . _ * 44 * , 184008 ( 2011 ) .
abate , j. a. , kimble , h. j. & mandel , l. photon statistics of a dye laser .
a _ * 14 * , 788 ( 1976 ) .
balakrishnan , n. , johnson , n. l. j. & samuel , k. a note on relationships between moments , central moments and cumulants from multivariate distributions .
_ statistics & probability letters _ * 39 * , 49 ( 1998 ) .
stanojevic , j. _ et al .
_ generating non - gaussian states using collisions between rydberg polaritons . _
a _ * 86 * 021403 ( 2012 ) . | ultra - cold gases excited to strongly interacting rydberg states are a promising system for quantum simulations of many - body systems @xcite . for off - resonant excitation of such systems in the dissipative regime ,
highly correlated many - body states exhibiting , among other characteristics , intermittency and multi - modal counting distributions are expected to be created @xcite .
so far , experiments with rydberg atoms have been carried out in the resonant , non - dissipative regime @xcite . here
we realize a dissipative gas of rubidium rydberg atoms and measure its full counting statistics for both resonant and off - resonant excitation .
we find strongly bimodal counting distributions in the off - resonant regime that are compatible with intermittency due to the coexistence of dynamical phases .
moreover , we measure the phase diagram of the system and find good agreement with recent theoretical predictions @xcite .
our results pave the way towards detailed studies of many - body effects in rydberg gases .
ultra - cold atoms excited to high - lying rydberg states have been shown in recent years to be promising candidates for experimental implementations of quantum computation and quantum simulation @xcite .
the strong and controllable interactions between rydberg atoms mean that fast two - qubit quantum gates and models of many - body hamiltonians can , in principle , be efficiently realized
. as a further important ingredient , dissipation has been shown to lead to novel phases and to facilitate , under certain conditions , the creation of pure and coherent many - body states @xcite .
also , a recent proposal for a quantum simulator based on rydberg atoms relies on dissipation @xcite .
+ so far , rydberg excitations in cold gases have been studied almost exclusively in the resonant , non - dissipative regime . in those experiments ,
the strong interactions between rydberg atoms manifest themselves either as spatial correlations compatible with a radius of blockade around an excited atom @xcite or through a reduction of fluctuations leading to sub - poissonian statistics @xcite . in this work
we show that for off - resonant excitation , the behaviour of the system depends strongly on the detuning and the sign of the rydberg - rydberg interaction .
we characterize the properties of our system through the full counting statistics of the excitation events , similarly to the methods recently used in condensed matter physics to unveil correlations in electronic transport processes @xcite .
when the detuning and the interaction have the same sign , excitation of pairs of rydberg atoms can occur , leading to an asymmetry of the excitation lineshape as a function of detuning . in the dissipative regime , we observe strongly bimodal counting distributions in the full counting statistics @xcite , indicating phase coexistence and intermittency in the system @xcite .
+ figure 1 summarizes the experimental method and typical results of the full counting statistics .
our experiments are performed using 87-rubidium atoms in magneto - optical traps ( mots ) that are excited to @xmath0 rydberg states using a two - step excitation process with detuning @xmath1 from resonance and two - photon rabi frequencies of up to @xmath2 @xcite ( see methods for details ) .
after the excitation pulse the rydberg atoms are field ionized and detected using a channeltron charge multiplier , whose output is recorded on an oscilloscope .
three different excitation regimes can be distinguished .
away from resonance , two rydberg atoms at a distance @xmath3 from each other can undergo a pair excitation if their van - der - waals interaction , characterized by the @xmath4 coefficient , matches twice the energy mismatch of the excitation , i.e. @xmath5 , as shown in fig .
1d . for the van - der - waals coefficient of the @xmath0 rydberg state in 87-rb ( @xmath6 ) and a typical detuning of around @xmath7 ,
this corresponds to a resonant distance @xmath8 ( where the range in @xmath3 is due to the finite laser linewidth of about @xmath9 ) .
this resonant condition is the exact opposite of the blockade effect ( fig .
1 c ) , where the interaction leads to the _ suppression _ of excitations , allowing at most one single excitation within a blockade radius ( which is collectively shared by all the atoms in the blockade volume ) .
we can estimate the timescale on which off - resonant excitations become important by comparing the resonant collective rabi frequency @xmath10 ( where @xmath11 is the number of atoms inside a blockade volume ) @xcite with the off - resonant two - photon rabi frequency @xmath12 expected for the two - photon excitation of two rydberg atoms with an intermediate singly - excited state detuned by @xmath1 from resonance . for typical values of our experiment ( @xmath13 , @xmath14 )
we find @xmath15 and @xmath16 at a detuning @xmath17 , leading to an off - resonant excitation timescale of around @xmath18 ( subsequent excitations mediated by already excited atoms can occur on a shorter timescale ) .
finally , for a detuning with opposite sign to that of the van - der - waals interaction , neither single - particle nor pair excitations are resonant , leading to a strong overall suppression of the excitation probability ( fig .
1b ) . + these three excitation regimes are summarized in fig .
1e , where the mean number of rydberg excitations is plotted as a function of detuning for two different excitation durations .
for the @xmath19 pulse off - resonant excitations are negligible and hence the lineshape is symmetric .
if the pulse duration is increased to @xmath20 , however , the excitation time is an appreciable fraction of the off - resonant excitation timescale , and off - resonant excitations become visible .
this leads to a lineshape that is clearly asymmetric with respect to zero detuning .
further evidence for the different excitation regimes is found in the full counting statistics of the rydberg excitations , as represented in fig .
1f - h by the histograms of the counting distribution .
these histograms clearly reveal the qualitative differences between on- and off - resonant excitations : on resonance , the distribution is roughly poissonian ( it becomes sub - poissonian for even longer excitation durations , for which the system enters the fully blockaded regime , as seen in fig .
3b ) .
by contrast , for positive detuning the distribution becomes multi - modal with two dominant features : one close to @xmath21 excitations and the other centered around a mean value of @xmath22 . for negative detuning , the histogram confirms the expected strong suppression of excitations .
+ the main features of the experimentally observed lineshape and counting distributions are reproduced by a numerical simulation based on a dicke - type model , as explained in the supplementary information .
the main results of that simulation are shown in fig .
2a - d .
whereas the model is not expected to give quantitative agreement with the experiment ( as it does not , for instance , take into account the spatial inhomogeneity of the cloud , the residual velocity of the atoms or decay from the rydberg state ) , for reasonable choices of the parameters in the simulation the qualitative agreement with the experimental data of fig .
1 is rather good .
+ we now turn to the dissipative excitation regime in which several excitation - spontaneous emission cycles occur . even though our experimental system does not allow us to work in the fully coherent regime ( due to limitations regarding the laser linewidth , and hence coherence time , and the maximum rabi frequency ) , the @xmath20 excitation duration used for the experiments shown in fig .
1 is an order of magnitude shorter than the lifetime of the @xmath0 rydberg state of around @xmath23 .
a comparison of the counting distributions for different excitation durations ( in which the rabi frequency was adjusted in order to keep the mean value constant , allowing a direct comparison between the histograms ) , as shown in fig .
2e , suggests that the counting distribution becomes more strongly bimodal for longer excitation durations , indicating that dissipation favors the appearance of bimodality ( as theoretically predicted in @xcite ) .
this is confirmed by the increase in the bimodality parameter ( see methods ) as a function of the duration ( again for fixed mean rydberg number ) .
+ acquiring the complete counting distribution allows us to gain more insight into the properties of our system in the dissipative regime beyond the behaviour of the mean and the variance @xcite .
we have characterized the full counting statistics on resonance and for positive detuning over a range of rabi frequencies by analyzing the second , third and fourth central moments @xmath24 , @xmath25 and @xmath26 as well as the associated normalized quantities ( mandel @xmath27-factor , skewness , binder cumulant and bimodality coefficient ) as a function of the mean number .
fig .
3 illustrates the main results of this analysis .
figs . 3a and 3b show the off- and on - resonant counting distributions for two mean values ( the same in both graphs ) , illustrating the bimodality for off - resonant excitation and the sub - poissonian character of the distribution on resonance .
this difference is clearly seen in the dependence of @xmath24 and the mandel @xmath27-factor on the mean number : whereas for on - resonant excitation those quantities are consistent with poissonian distributions for small mean numbers , they become increasingly mono - modal ( i.e. , strongly sub - poissonian ) for larger numbers as the system enters the fully blockaded regime . by contrast , in the off - resonant case @xmath24 initially _ increases _ with the mean , reaching a peak at about half the maximum number of excitations ( the mandel @xmath27-factor also exhibits a maximum , which is shifted to smaller numbers due to the normalization by the mean ) . as can be seen in fig .
3c - h , the results of the off - resonant case are in qualitative agreement with a simple bimodal model . in particular , we find that the bimodality coefficient ( fig .
3h ) of the off - resonant counting distributions is consistently higher ( at @xmath28 ) than for the resonant case ( around @xmath29 ) over the entire range of the mean number , emphasizing the qualitative difference between the two regimes .
it is also obvious , however , that while the agreement is good for @xmath24 , the deviation increases for the higher central moments .
this demonstrates the usefulness of the full counting statistics , which allows comparisons with theoretical predictions that are far more sensitive than the mean and standard deviation alone .
+ the results shown in fig .
3 can be interpreted in terms of a dynamical phase transition between a paramagnetic and an antiferromagnetic phase in a dissipative ising model with a transverse field , as shown recently in @xcite .
although in those works the atoms are assumed to be arranged in a crystalline structure ( created by an optical lattice ) with regular spacing , the distance - selective resonance mechanism described above means that in our experiment the rydberg excitations should arrange themselves in a regular array with a spacing that depends on the detuning , as suggested theoretically in @xcite . in fig
4 we show a phase diagram for our system as a function of the rabi frequency and the detuning . the mean rydberg number as a function of the rabi frequency exhibits a smooth crossover between @xmath21 excitations and a maximum number of around @xmath30 excitations ( see also fig . 3a ) , where the position of the crossover depends on the detuning .
this is expected from the analogy with an ising spin system @xcite , where the critical value of the transverse field ( which corresponds to the rabi frequency in our system ) increases with increasing ising interaction ( corresponding to the detuning in our case ) .
moreover , we observe a distinct increase in the mandel @xmath27-factor in the transition region , which is compatible with the intermittent behaviour of the system theoretically predicted in @xcite ( and with our observation of bimodal counting statistics in that region ) , whereby in the transition region the active and inactive phases of the system coexist . in order to prove that interpretation directly , however , it will be necessary to observe the time evolution of a single experimental realization , e.g. , through the observation of photons emitted during the decay process @xcite .
+ in conclusion , we have analyzed resonant - and off - resonant rydberg excitations in a cold rubidium gas in the dissipative regime through full counting statistics .
we have shown that the full counting statistics reveals characteristic features of the system that are not evident in the mean or standard deviation typically measured in such experiments .
our technique should be useful for the characterization of rydberg excitations in optical lattices @xcite in which , e.g. , the ising model can be realized , and in experiments with chirped excitation lasers aimed at the adiabatic creation of rydberg crystals @xcite .
more generally , the full counting statistics will be an important tool for unveiling many - body effects in rydberg excitations , for instance in quantum simulators . |
the localization model introduced by anderson @xcite incorporates two basic elements , namely the rules of quantum mechanics applied to a single - electron , tight - binding model hamiltonian , plus quenched disorder ( realized , e.g. , by assigning random self - energies to lattice sites ) .
its original purpose was to show the existence of a disorder - induced transition in three - dimensional systems , from metallic ( diffusive ) to insulating ( localized ) electronic behavior , upon increasing randomness . over the years
the model has turned out to exhibit a rich variety of physical aspects , many of them highlighted by the ( single parameter ) scaling theory of localization ( spst ) @xcite .
of particular interest here is the fact that , in zero magnetic field and in the absence of spin - orbit couplings , spst predicts insulating behavior , for any finite amount of disorder , in spatial dimensions @xmath1 and @xmath6 , though in the marginal case @xmath2 one has borderline phenomena such as weak localization .
interest in the anderson transition has been renewed by reports of metallic behavior in dilute two - dimensional electron - hole systems @xcite .
while phenomenological , percolation - based theories have been able to reproduce experimentally - observed trends in some detail @xcite , attempts to reconcile basic theoretical assumptions to experimental evidence have only met limited success so far .
for instance , numerical evidence has been produced @xcite against the idea that electron - electron interactions ( not included in scaling theory ) might play a role in driving the two - dimensional transition @xcite .
even when one confines oneself to the original anderson picture of non - interacting electrons in three - dimensional lattices , where the existence of a transition is not questioned , progress towards extracting reliable numerical estimates of critical quantities has been remarkably hard @xcite .
recently , systematic consideration of irrelevant variables and non - linear corrections to single - parameter scaling @xcite has helped produce results with a fairly reasonable claim to consistently narrow error bars .
it is thus of interest to reexamine the basic methods which have been used in the past 20 years , in conjunction with spst , to study the anderson localization problem .
a step in this direction has been given in ref .
, whose authors obtained wave functions via exact diagonalization , for both one - dimensional and finite , @xmath7 , two - dimensional systems . by averaging over randomness ,
they obtained probability distributions of wavefunction amplitudes on sites at varying distances from an arbitrary origin .
such distributions were compared to predictions from spst ; though agreement was good in @xmath1 , the two - dimensional results were in contradiction to the idea of a single localization length depending only on disorder intensity : instead , a clear logarithmic increase with distance , at fixed disorder , was found from their fits for that quantity . in ref .
, finite - size effects were avoided in @xmath2 by considering disorder strengths such that the corresponding localization length , as predicted by single - parameter theory , is @xmath8 ( see , e.g. , refs . ) , and using suitably large systems with @xmath9 . on the other hand ,
numerous studies of the anderson transition are set up on quasi - one dimensional geometries , for ease of application of transfer - matrix ( tm ) or recursive green s functions methods @xcite ; extrapolation to bulk behavior ( @xmath2 or @xmath10 as the case may be ) is then performed with help of finite - size scaling theory @xcite .
here we consider tm methods , applied both to strictly one - dimensional systems and to strips of a square lattice .
traditionally the tm approach has been used to calculate lyapunov exponents ( directly related to the localization length of spst ) @xcite , and quantities obtainable from such exponents , e.g. , conductances @xcite . in sec .
[ method ] we recall how wavefunction amplitudes may be estimated in the tm context , and illustrate our approach in the simple @xmath1 case by rederiving the corresponding distributions found in ref . . in sec .
[ twod ] , an analogous treatment is developed for strips of a two - dimensional square lattice , and numerical results are displayed and discussed .
conclusions and final remarks are given in sec .
we consider the site - disordered anderson model , for which the tight - binding hamiltonian is written as @xmath11 where the site self - energies @xmath12 are independent , identically distributed random variables obeying a specified distribution , @xmath13 denotes nearest - neighbor sites on a regular lattice , and the energy scale is set by the hopping matrix element , @xmath14 .
disorder intensity is given by the width @xmath15 of the self - energy probability distribution , taken here as rectangular ( same as in ref . ):
@xmath16 in the tm approach @xcite , one considers the hamiltonian eq .
( [ eq : ham ] ) on a quasi - one dimensional @xmath17 system , @xmath18 . denoting by @xmath19 the successive cross - sections , and @xmath20 the respective positions of sites within each cross - section ,
an electronic wave function at energy @xmath21 is given in terms of its local amplitudes , @xmath22 , and tight - binding orbitals @xmath23 , as : @xmath24 with a corresponding change of notation , eq .
( [ eq : ham ] ) reads : @xmath25 where @xmath26 stands for nearest - neighbor pairs . applying eq .
( [ eq : h(ik ) ] ) to eq .
( [ eq : psi(e ) ] ) gives the recursion relation @xmath27 where @xmath28 denotes nearest neighbors of @xmath29 within the same cross - section ( the @xmath21-dependence is omitted for clarity ) .
in matricial form , @xmath30 where , considering , e.g. , periodic boundary conditions across the @xmath31 transverse directions @xcite , @xmath32 the @xmath33 matrix @xmath34 is symplectic , that is , its eigenvalues occur in pairs @xmath35 . as explained at length in refs . , the matrix product @xmath36 gives rise to the eigenvalues @xmath37 , where the @xmath38 are the lyapunov characteristic exponents ( lce ) for the problem , setting the asymptotic divergence of the corresponding eigenvectors @xmath39 .
being a product of symplectic matrices , @xmath40 also has this property , therefore the lce occur in symmetric pairs @xmath41 .
attention usually concentrates on the lce of smallest modulus , @xmath42 , whose inverse gives the longest decay length ( identified with the localization length of the anderson problem ) . to see the meaning of the eigenvectors @xmath39 , recall that the physically acceptable ( non - diverging ) wave function is the one associated with the _ negative _ lce of smallest modulus , @xmath43 @xcite ; the corresponding tight - binding amplitudes @xmath44 will give information on the shape of the electronic wave function of interest ( because of the symplectic character of the tm , one might equally consider the _
inverse _ of the amplitudes associated to @xmath42 ; however , in practice the amount of calculational effort is the same either way ) .
little use appears to have been made of this property in the context of tm studies of anderson localization , except for a calculation of lateral transport properties in layered media @xcite .
it is important to recall that the site amplitudes are not the directly relevant quantities in the tm approach ; instead , in the quasi - one dimensional systems used here the wavefunction decay rate must be defined by comparing the moduli of suitable vectors , each with @xmath45 components @xcite .
we now take strictly one - dimensional systems , and illustrate how the above ideas work .
the matrices @xmath46 are @xmath47 , and the relevant lce is @xmath48 ; the corresponding wavefunction amplitude at site @xmath49 is @xmath50 . starting with an arbitrary pair of states at neighboring sites ,
say @xmath51 , we first iterate eq .
( [ eq : matrix ] ) a number @xmath52 of times , taking care to orthonormalize the resulting vectors every @xmath53 steps ( typically , @xmath54 ; we have used @xmath55 , but other authors have used @xmath56 apparently without noticeable deterioration of results @xcite ) . with such initialization the starting vectors are rotated in hilbert space towards the asymptotic direction of the eigenvectors of @xmath40 .
having done this , we rename the current site as the origin . recalling from eq .
( [ eq : matrix ] ) that the vectors being iterated involve both @xmath57 and @xmath58 , one sees that the appropriate quantities to keep track of are the @xmath59 ^ 2+[a_{k+1}^{(2)}]^2\}^{1/2}$ ] .
one might visualize the process as follows : starting from the pair of site amplitudes @xmath60 , one iteratively obtains the pair @xmath61 and so on , until ( after @xmath62 iterations of @xmath63 ) one gets the pair @xmath64 .
this latter is legitimately said to be at a distance @xmath62 from the origin , that is , from the original pair of sites .
we then start to accumulate the products of successive amplitudes @xmath65 at each @xmath53 steps ( _ after _ orthogonalization , but _ before _
normalization ) . at distance @xmath62 from the new origin ,
the ( relative ) wavefunction amplitude is given by @xmath66 where the notation of ref .
is used for ease of comparison , @xmath67 ^ 2+[\psi_{k+1}]^2\}^{1/2}$ ] , and the fact that we always make @xmath55 has been taken into account . in order to generate statistics of the @xmath68 for a set of distances @xmath69 , one iterates the tm for @xmath70 steps , collecting data at the specified points ; to collect the next sample , it suffices to keep iterating for another @xmath71 steps , with no need to reinitialize the wave functions , and so on .
after a total of @xmath72 iterations of the tm , one has @xmath73 samples of the @xmath68 for each distance of interest , the corresponding histograms @xmath74 for @xmath75 , disorder strength @xmath76 , @xmath77 , @xmath78 , and @xmath79 , and @xmath80 are shown in fig .
[ fig : fig1 ] .
these values of the parameters were chosen in order to enable comparison with the exact diagonalization results displayed in figure 1 of ref . .
indeed , one finds excellent visual agreement between the respective data sets .
numerical analysis of the first three moments of our histograms shows that : ( i ) for all three distances , they are well - fitted by gaussians of the form @xmath81\ , \label{eq : gfit1d}\ ] ] with @xmath82 , @xmath83 , which are in rather good agreement with the expected values from spst @xcite in the limit @xmath84 , respectively @xmath85 , @xmath86 ; and ( ii ) the dimensionless skewness @xmath4 , defined as @xcite : @xmath87 for a distribution with mean @xmath88 and dispersion @xmath89 , has the following values : @xmath90 , @xmath91 , @xmath92 respectively for @xmath77 , @xmath78 , @xmath79 . point ( ii ) is further indication that the wavefunction amplitude distribution indeed approaches a log - normal shape ( zero skewness ) , but only as @xmath93 . the approximate dependence on @xmath62 may be inferred as @xmath94 , from the 3 data just quoted .
we now extend our approach to strips of a square lattice . for a strip of width @xmath95 , the matrices @xmath46 are @xmath96 , and
the relevant lce is @xmath97 ; the corresponding wavefunction amplitudes at column @xmath49 are @xmath98 , @xmath99 . taking into account the normalization of eq .
( [ eq : matrix ] ) , the appropriate decay factor here is @xmath100
^ 2\,\right)^{1/2 } \ \ ( d=2 ) .
\label{eq : a(r)2d}\ ] ] it must be stressed that eq .
( [ eq : a(r)2d ] ) is not meant to imply an averaging process over site amplitudes @xmath101 ; as remarked above , these @xmath102 quantities are not the directly relevant ones . instead
, they give the modulus of the eigenvector associated to the negative lce of smallest absolute value @xcite , whose decay is to be followed .
we have considered strips of even widths @xmath103 sites and periodic boundary conditions across , and taken the disorder intensity @xmath104 , in order to make contact with analogous results in ref . .
for this value of @xmath15 spst predicts the localization length to be @xmath105 @xcite .
spst , together with finite - size scaling @xcite would imply that @xmath106 in order to infer the @xmath2 behavior , one must consider the regime @xmath62 , @xmath107 , @xmath108 @xcite . although tm methods make it easy to explore long distances ( @xmath109 ) along the `` infinite '' direction , this fact is not directly relevant here , as the corresponding regime would be one where strictly one - dimensional features emerge . in the following
, we shall restrict ourselves to @xmath110 and @xmath111 ; according to eq .
( [ eq : spst2d ] ) , for each value of @xmath112 one should then be able to collapse all distributions against @xmath113 .
again , we have examined the first three moments of the distributions thus generated . using only the first two , we have fitted data to gaussians in the manner of eq .
( [ eq : gfit1d ] ) , for which the effective localization lengths @xmath114 are displayed in fig .
[ fig : fig2 ] .
it can be seen that , for given @xmath95 , the @xmath115 are essentially the same both for @xmath116 and @xmath95 .
this shows that crude finite - size distortions do not play a role for the ranges of @xmath62 and @xmath95 used . on the other hand , similarly to the findings of ref .
and against the prediction of spst , there is no single value of @xmath115 to fit all distributions ; instead , it grows with increasing @xmath62 .
however , our result differs from that of ref .
, in that the dependence of @xmath115 is clearly not linear in @xmath117 .
this should not be seen as a direct contradiction , as the quantities under study are not identical ( as was the case in @xmath1 ): though they represent the same physical phenomenon of wavefunction decay , they do so in rather different geometries . as regards the width of distribution @xmath118 ( again taking eq .
( [ eq : gfit1d ] ) as a starting point ) our results , displayed in the inset of fig .
[ fig : fig2 ] , exhibit numerical values not unlike those found in ref . , in the sense of being consistently smaller than the spst prediction @xmath85 , but with the same order of magnitude .
one might expect that , for larger @xmath62 , @xmath95 the decreasing trend observed for @xmath119 would stabilize close to @xmath120 quoted in ref . .
this , however , we have no means to ascertain at present .
we have not been able to fit the full range of data by a single power of either @xmath95 or @xmath117 ; assuming , e.g. , @xmath121 , the best result from a nonlinear least - squares fit gives @xmath122 , corresponding to the dashed line in fig .
[ fig : fig2 ] .
it is evident , from the figure , that the trend for large @xmath95 is towards an even slower variation . at this point , one might speculate that @xmath115 , as given by the gaussian fits , could eventually saturate for larger @xmath95 , at a value which might even be close to the spst prediction .
however , we shall now show that the gaussian fits themselves become increasingly unable to reflect the properties of the wavefunction amplitude distributions .
indeed , we have found that the skewness of distributions is negative , and _
increases _ in absolute value as @xmath62 , @xmath95 grow . as an example ,
[ fig : fig3 ] shows the raw data for @xmath123 , @xmath124 , together with the corresponding gaussian fit of eq .
( [ eq : gfit1d ] ) , obtained using the first and second moments of the distribution .
the effect of negative skewness is apparent in that the gaussian approximation overshoots the data for large @xmath125 ( i.e. , predicts a _ small _ amplitude to occur _ more frequently _ than observed in fact ) , and undershoots for small @xmath125 ( predicts a _ large _ amplitude to occur _ less frequently _ than observed ) .
skewness data for the ranges of @xmath62 and @xmath95 used here are displayed in fig .
[ fig : fig4 ] , together with fits of single - power forms , @xmath126 , for the subsets corresponding respectively to @xmath127 ( dashed line , @xmath128 ) and @xmath116 ( full line , @xmath129 ) . despite the large amount of scatter , the increasing trend against growing @xmath62 , @xmath95 is unmistakably present .
this means , in turn , that for larger and larger systems the distributions become ever less amenable to fitting by gaussians , as predicted by spst .
this would remain true even if a hypothetical saturation should occur for values of @xmath62 and @xmath95 larger than those investigated here .
recall , from eq .
( [ eq : skewdef ] ) , that skewness is invariant under the usual histogram - collapse rescaling @xcite , @xmath130 , where the shifted variable is @xmath131/\sqrt{\sigma r/\lambda ( r)}$ ] . therefore , this is a legitimate extra parameter to characterize the distributions .
similar results were found experimentally , for the conductance distribution in quasi - one dimensional gold wires @xcite .
negative skewness of wavefunction amplitude distributions works in the same way as ( in the limited context of gaussian fits ) does the finding that @xmath132 increases with @xmath62 : both contribute to a slower decay of electronic wave functions , compared with the constant-@xmath115 , zero - skewness , spst picture .
of course , the evidence just presented is not enough to argue that there must be a localization - delocalization transition in @xmath2 ; the idea that this is the borderline dimensionality , as predicted by spst , most likely holds true .
nonetheless , we have shown robust evidence for deviations from spst in @xmath2 , whose consequences still have to be worked out in full ,
we have made use of suitable properties of the vectors which are iterated in the tm approach to anderson localization @xcite , in order to generate the statistical distribution of electronic wavefunction amplitudes at sites of @xmath0 disordered systems .
we have considered @xmath1 ( for which our approach is shown to reproduce the exact diagonalization results of ref . ) , and @xmath2 . in the latter case , since the @xmath133 geometry of our systems differs from that ( @xmath7 ) of ref .
, a perfect match is not to be expected ; however , some basic physical properties are found to hold for both cases . in particular , attempted fits of gaussian ( log - normal ) forms to the wavefunction amplitude distributions result in effective localization lengths growing with distance , contrary to the spst prediction .
we have gone further , and shown that the distributions possess a negative skewness , which is invariant under the usual histogram - collapse rescaling , and increases with distance ( at least for the range of parameters used in our study ) .
such deviations from the expected behavior are evidence of slower decay of electronic wavefunctions than predicted by spst ; it still must be worked out whether or not some phenomena specific to @xmath2 , such as weak localization , or the recently - observed metal - insulator transition in dilute two - dimensional dilute electron - hole systems @xcite , carry the fingerprints of the anomalies reported here .
we expect that the present results may motivate further work along these lines .
p. w. anderson , phys .
rev . * 109 * , 1492 ( 1958 ) .
e. abrahams , p. w. anderson , d. c. licciardello , and t. v. ramakrishnan , , 673 ( 1979 ) .
a. mackinnon and b. kramer , rep .
* 56 * , 1469 ( 1993 ) . for a review ,
see e. abrahams , s. v. kravchenko , and m. p. sarachik , , 251 ( 2001 ) .
y. meir , , 3506 ( 1999 ) ; , 16@xmath134470 ( 2000 ) .
see , e.g. , r. berkovits , j. w. kantelhardt , y. avishai , s. havlin , and a. bunde , , 082192 ( 2001 ) and references therein .
pichard and g. sarma , j. phys .
c * 14 * , l127 ( 1981 ) ; * 14 * , l617 ( 1981 ) .
a. mackinnon and b. kramer , , 1546 ( 1981 ) .
a. mackinnon and b. kramer , z. phys .
b : condens . matter * 53 * , 1 ( 1983 ) .
b. bulka , b. kramer , and a. mackinnon , z. phys .
b condens .
matter * 60 * , 13 ( 1985 ) .
b. bulka , m. schreiber , and b. kramer , z. phys .
b condens .
matter * 66 * , 21 ( 1987 ) .
chang , j. d. bauer , and j. l. skinner , , 8973 ( 1990 ) .
a. mckinnon , j. phys . : condens .
matter * 6 * , 2511 ( 1994 ) .
j. l. cardy , _ scaling and renormalization in statistical physics _
( cambridge university press , cambridge , 1996 ) , chap .
k. slevin and t. ohtsuki , , 382 ( 1999 ) .
t. ohtsuki , k. slevin , and t. kawarabayashi , ann . phys .
( leipzig ) * 8 * , 655 ( 1999 ) .
k. slevin , t. ohtsuki , and t. kawarabayashi , , 3915 ( 2000 ) .
k. slevin , p. markos , and t. ohtsuki , , 3594 ( 2001 ) .
s. l. a. de queiroz , , 214202 ( 2001 ) . j. w. kantelhardt and a. bunde , , 035118 ( 2002 ) . m. n. barber , in _ phase transitions and critical phenomena _ , edited by c. domb and j. l. lebowitz ( academic , new york , 1983 ) , vol .
a. crisanti , g. paladin , and a. vulpiani , in _ products of random matrices in statistical physics _ ,
springer series in solid state sciences vol .
104 , edited by helmut k. lotsch ( springer , berlin , 1993 ) .
pichard , j. phys .
c * 19 * , 1519 ( 1986 ) .
pichard and g. andr , europhys . lett .
* 2 * , 477 ( 1986 ) .
f. delyon , y. e. lvy and b. souillard , , 618 ( 1985 ) ; b. souillard , in _ chance and matter , les houches session xlvi _ , edited by j. souletie , j. vannimenus and r. stora ( amsterdam : north - holland , 1987 ) , pg .
chan and z .- q .
zhang , , 36 ( 1997 ) .
m. kappus and f. wegner , z. phys .
b : condens . matter * 45 * , 15 ( 1981 ) . w. press , b. flannery , s. teukolsky , and w. vetterling , _ numerical recipes in fortran , the art of scientific computing _
( cambridge university press , cambridge , 1994 ) , sec . 14.1 .
p. mohanty and r. a. webb , , 146601 ( 2002 ) . | we show how to use properties of the vectors which are iterated in the transfer - matrix approach to anderson localization , in order to generate the statistical distribution of electronic wavefunction amplitudes at arbitrary distances from the origin of @xmath0 disordered systems . for @xmath1 our approach
is shown to reproduce exact diagonalization results available in the literature . in @xmath2 , where strips of width @xmath3 sites were used , attempted fits of gaussian ( log - normal ) forms to the wavefunction amplitude distributions result in effective localization lengths growing with distance , contrary to the prediction from single - parameter scaling theory .
we also show that the distributions possess a negative skewness @xmath4 , which is invariant under the usual histogram - collapse rescaling , and whose absolute value increases with distance .
we find @xmath5 for the range of parameters used in our study . |
there are many classes of binary systems which can not be understood without a phase of significant mass and angular momentum loss . among
those classes are interacting binary systems containing white dwarf or neutron star components .
common envelope evolution has been identified as a means by which the interactions between the two progenitor components lead to the ejection of a large fraction of the mass of the system and to the shrinkage of the binary orbit ( see paczynski 1976 ) . provided
that the common envelope is successfully ejected , a long - period system ( @xmath9 years ) can be transformed into a short - period system ( @xmath9 days ) consisting of the remnant core of the red giant and its companion .
a binary system can evolve into the common envelope stage whenever the red giant progenitor of the compact component enters a phase where it is not corotating with respect to the orbit at the onset of or during mass transfer to its companion .
this may occur when mass transfer from the more massive star to its companion is unstable , as occurs when the roche lobe contracts faster than its radius , especially in the case of a giant with a deep convective envelope ( see paczynski & sienkiewicz 1972 ; webbink 1979 ) .
another evolutionary path leading to the common envelope stage involves a tidal instability ( darwin 1879 ; counselman 1973 ; kopal 1978 ; lai , rasio , & shapiro 1993 , 1994 ) , which occurs when a system reaches a minimum value of the total angular momentum for a synchronized close binary system . evolution to this stage inevitably leads to the infall of one component into its more massive companion . as a result of the development of powerful numerical algorithms and the advancement of computer technology ,
progress has been made in modeling common envelope interactions from the slow initial phase to the rapid phase of orbital shrinkage .
many of the simplifying approximations of the past regarding the input physics , as well as the dimensionality of the problem , have been relaxed . in particular , the multidimensional studies of bodenheimer & taam ( 1984 ) , livio & soker ( 1988 ) , and taam & bodenheimer ( 1989 ) revealed that matter is preferentially ejected at higher than the escape speed along the equatorial plane of the binary system .
this theoretical result received observational support from the elliptical or butterfly - shaped appearance of planetary nebulae with binary nuclei ( bond & livio 1990 ) .
that is , the morphology of such nebulae are consistent with the density contrast between the pole and equator seen in common envelope simulations used in interacting winds models ( kwok 1982 ; soker & livio 1989 ; frank et al .
1993 ; mellema & frank 1995 ; livio 1995 ) .
this nonspherical ejection follows from the angular distribution of the energy and angular momentum deposited in the common envelope by the orbital motion of the two components .
population synthesis calculations by a number of investigators ( see , for example , tutukov and yungelson 1979 ; iben & tutukov 1984 ; van den heuvel 1987 ; de kool 1992 ; iben & livio 1993 ; yungelson , tutukov , & livio 1993 ; han , podsiadlowski , & eggleton 1995 ; kalogera & webbink 1996 ; de kool 1996 ) have used approximations to the common envelope phase to predict the properties of systems which survive .
such calculations make use of energetic arguments , and the results are sensitive to the adopted value of the efficiency of the common envelope phase ( taken to be equal to the ratio of the binding energy of the ejected mass to the energy lost from the binary orbit ) .
although the results of the early numerical studies of taam , bodenheimer , & ostriker ( 1978 ) , meyer & meyer - hofmeister ( 1979 ) , delgado ( 1980 ) , bodenheimer & taam ( 1984 ) , livio & soker ( 1984 , 1988 ) , and taam & bodenheimer ( 1989 , 1991 ) were instructive , the important details of the outcome of the common envelope phase were still unclear , since understanding of the mass ejection process was affected by the approximations ( such as using smaller numbers of spatial dimensions , or imposed descriptions for the orbital energy dissipation ) needed to make the problem tractable .
thus , the production of short - period binary systems containing compact objects remained to be clarified ( even if sufficient energy was lost from the orbit to unbind the common envelope ) since the results of detailed calculations have not been taken into account in the population synthesis studies in determining whether the two stars would stop spiraling toward each other ( see taam & bodenheimer 1991 ) .
in this paper , we investigate the details of the hydrodynamics of the common envelope phase in the vicinity of the double core at high spatial resolution in three dimensions in order to understand the termination of the spiral - in phase .
previous three - dimensional studies ( de kool 1987 ; terman et al . 1994 , 1995 ; terman & taam 1996 ; rasio & livio 1996 ) have relied on the smoothed particle hydrodynamics technique ( sph ; e.g. , lucy 1977 ; gingold & monaghan 1977 ; monaghan 1985 , 1992 ) .
the numerical results of such studies indicated that mass is ejected from the common envelope , although the poor spatial resolution of the simulations made it unclear whether the companion would spiral into the red giant core .
although the sph technique can model the global features of the evolution during earlier stages , our experience indicates that the numerical resolution is not entirely adequate to model the region of steep density gradients above the red giant core during the terminal phase of evolution .
in addition , the large number of time steps required to follow the ejection of a significant fraction of the envelope using higher resolution is computationally prohibitive . to adequately resolve the region close around the double core
, we use a finite differencing numerical technique .
in contrast to the study of livio & soker ( 1988 ) , we use an eulerian nested grid technique similar to that used in yorke et al .
( 1995 ) , but extended to three dimensions .
we report on the results of simulations of the common envelope phase for binaries consisting of an intermediate - mass asymptotic giant branch star of either 3 or 5 @xmath10 with a main sequence companion of 0.4 or 0.6 @xmath10 . in the following section we describe the numerical methods used , the assumptions underlying the calculations , and the construction of the initial models .
the detailed evolution of a binary from the initial spiral - in phase through the late phase of envelope ejection is presented in 3 , as are comparisons with our other simulations , and a discussion of physics affecting the system beyond the scope of our simulations .
the implications of the simulations are discussed in the final section .
for all of the simulations in this paper , a 3-d grid - based hydrodynamics code is used .
a nested grid technique is employed to provide higher spatial resolution in the inner regions of the common envelope , with dynamical coupling between grids based on the technique of berger & oliger ( 1984 ) and berger & colella ( 1989 ) , as described in yorke , bodenheimer , & laughlin ( 1993 ) and ruffert ( 1993 ) .
we have modified a version of the code used by burkert & bodenheimer ( 1993 ) , which they used to follow the isothermal gravitational collapse of a protostar . in our simulations
, the main grid has @xmath11 cubical zones , while the nested subgrids have @xmath12 cubical zones , where the short dimension is perpendicular to the orbital plane .
two subgrids were centered within the main grid , and were kept motionless with respect to it .
the first subgrid was a factor of 4 smaller in the @xmath13 and @xmath14 directions , and the second subgrid was a factor of 2 smaller than the first , so that the zone size on the finest grid was @xmath15 cm in all but one of the simulations ( for which it was twice as large ) .
the total mass , energy , and angular momentum of the gas lost from the main grid were followed . in order to simulate realistic interactions between a giant star and a main sequence companion
, the code also follows the motions of two collisionless particles one for the core of the red giant star , and one for the companion .
gravitational interactions between the collisionless particles and the gas were smoothed in a manner suggested by ruffert ( 1993 ) : @xmath16 where @xmath17 is the width of a zone on the innermost subgrid , and @xmath18 . here
@xmath19 is the distance between the gas and the particle of mass , @xmath20 .
the gravitational potential of the gas was computed separately on the main grid and subgrids via fast fourier transforms .
boundary values for the subgrid computations were taken from the smallest grid that encompassed it .
timesteps for each grid were chosen according to a courant - friedrichs - lewy condition on that grid : @xmath21 where @xmath22 is the courant factor , and @xmath23 is the sound speed . on the innermost subgrid , the timestep can be reduced to @xmath24 if this is smaller than the timestep derived from the gas , where @xmath25 and @xmath26 are the speeds of the two collisionless particles .
changes to the courant number were made as tests , but these had no effect on the results .
core particles were stepped in position using a four - point runge - kutta integration scheme each time the gas distribution was recalculated on the innermost subgrid . when the cores are at their closest approach , the orbits only pass through a handful of zones , so that with the courant - type timestep , the final orbit would be poorly integrated . to improve on this , the gas timestep was increasingly subdivided to ensure that the core positions are computed at least 25 times during each orbit . because the original version of the code was intended for use in computing isothermal collapse , the code was modified to explicitly treat the internal energy of the gas .
the equation of state was chosen to be a combination of ideal gas and radiation pressure ( the regions where the gas is affected by electron degeneracy are sufficiently compact that they can be assumed to be restricted to the point masses ) .
internal energy density was advected in exactly the same manner as mass density , and compressional heating of the gas was also included .
shocks in the gas were handled with an artificial viscosity of the type introduced by von neumann and richtmeyer ( 1950 ) with a constant shock smoothing length of @xmath27 , where @xmath28 is the zone size in grid or subgrid @xmath29 .
advection of density , internal energy , and momentum across zone boundaries was handled with a monotonic transport formalism ( van leer 1977 ) , which provides second - order spatial accuracy , and quasi - second order accuracy in time .
a number of additional tests were conducted in order to check the accuracy of the code for common envelope simulations .
first , we conducted a standard shock tube run ( sod 1978 ) , verifying that the numerical results matched the analytical solution for adiabatic flow .
the results were substantially similar to those of burkert & bodenheimer ( 1993 ) .
we verified that the shocks were not disturbed by the presence or absence of subgrids in the calculation .
secondly , several trial common - envelope runs were conducted to check the accuracy of various aspects of the code .
the most important effects involved the smoothing length used in gravitational interactions between core particles and the gas .
the value of the smoothing length @xmath30 adopted is a minimum acceptable value lower values resulted in unacceptable non - conservation of total energy . because of the importance of understanding the orbital evolution of the core particles
after the majority of the gas has been forced away during the infall of the companion , the smoothing length was reduced as much as possible .
the same gravitational smoothing length was used in all levels of each simulation since it was found that significant non - conservation of total energy occurred if the point mass came near the edge of a subgrid during simulations in which the smoothing length was made equal to @xmath31 times the size of a zone in the grid which the particle was situated .
the degree of physical precision of our calculations depends on the resolution of the gas in the regions of strongest interaction ( the number of subgrids used ) and on the accuracy of the force calculation near the core particles ( the core - gas smoothing length ) .
the degree to which energy was conserved depended on the maximum strength of the core gravitational fields ( which was determined by the smoothing length ) . a similar effect
was found by ruffert ( 1993 ) , in that the density and temperature structure near the core was sensitive to the structure of the core potentials .
after a number of tests , it was found that for a given smoothing length , fewer subgrids resulted in slightly better conservation of energy .
this motivated our use of only two subgrids .
we will return to this topic in [ numerical ] , where we discuss the effects of resolution on the final orbital separation .
the initial gas distribution in the envelope of the red giant was derived from one - dimensional stellar models obtained from the code developed by eggleton ( 1971 , 1972 ) .
interpolation within a stellar model in the appropriate evolutionary state was used to obtain densities and internal energies for the zones .
a diffuse background gas filled the remainder of the grid , and was given a temperature to bring it into pressure equilibrium with the surface of the star .
the mass of the core particle for the giant was chosen to bring the total giant mass to that of the input stellar model .
the orbit of the binary was placed in the xy - plane .
we also forced the gas to be symmetric about this plane , so that at all times the core particles remained very close to z@xmath32 . in order to determine the sensitivities of the outcome of the common envelope evolution ,
we have varied three parameters : the companion mass , the red giant mass , and the fraction of synchronous rotation given to the envelope of the giant .
our baseline simulation , against which we will compare all of our runs , involved a @xmath33 red giant star ( with a @xmath34 core , and a total radius of @xmath35 cm ) , and a @xmath36 companion particle .
the companion was initially placed in a circular orbit at a distance of @xmath37 cm , with the giant in synchronous rotation with the companion . for the other simulations ,
these parameters were varied , and the alternate values used were : a @xmath38 companion ; a @xmath39 giant star ( with the same radius as the @xmath33 giant , but with a @xmath40 core ) ; and a non - rotating giant .
one additional simulation involving a more evolved @xmath39 giant with a radius of @xmath41 cm was run with a companion of @xmath42 orbiting at a distance of @xmath43 cm .
we have not started our baseline simulation in a relaxed - binary configuration as did rasio & livio ( 1996 ) , but we believe that the difference in the initial conditions is a minor consideration . in the rasio & livio s ( 1996 ) case , matter was barely overflowing the giant s roche lobe whereas in our initial configuration the mass transfer developed quickly .
this difference in initial conditions results in a sharper angle of impact for the companion star onto the giant s surface in our simulation .
in addition , the efficiency of mass ejection from the binary system is somewhat underestimated in the present study since the energy transfer from the orbit to the giant envelope to bring it to a roche - lobe filling configuration is already incorporated into the envelope in the initial state described by rasio & livio ( 1996 ) . however , this difference is not important for the final outcome since this aspect of the evolution affects only the near - surface gas , which is the least bound material .
hence , the energy difference is rather small in that sense .
most importantly , the general agreement in the final state of our simulations between the synchronously rotating and asynchronously rotating cases argues that the effects of the initial conditions are minor .
in particular , the evolution and timescales for the rapid dynamical phase and mass ejection phase are unaffected by these details of the initial configuration .
in this section we report on the numerical results of five simulations .
the initial parameters and results for each sequence are summarized in tables 1 and 2 respectively . for convenience ,
the results are presented in detail for the baseline simulation ( sequence 1 ) followed by descriptions of the differences that exist among the simulations . for sequence 1 ,
we consider the common envelope evolution of a binary consisting of a @xmath44 asymptotic giant branch ( agb ) star with a @xmath2 main sequence companion .
the two components are orbiting about their common center of mass in circular motion with a period of 1.2 years . in the earliest phase of the evolution , gas from the near surface of the giant star
is gravitationally torqued up , removing energy and angular momentum from the binary orbit in the process .
( the energy and angular momentum as a function of time are shown in figures [ erg1 ] and [ ang1 ] . )
we define this initial phase as the time interval between the beginning of the simulation ( a state approximating a realistic initial binary orbit ) and the time at which the separation of the point masses is equal to the initial radius of the giant . in our baseline simulation
, this phase lasts 160 days . during this initial phase
, there is a period of fairly heavy mass transfer from the giant to the companion .
the matter stream is dragged around behind the companion s direction of travel into an orbit until the stream impacts itself , after which it quickly fills the companion s potential well . by the time the companion reaches the former surface of the giant , the core and companion have transferred about one - third of their angular momentum to the gas .
very little mass has been unbound at this point though , and none has been lost .
in addition , the energies in the system have changed very little , as the gas surrounding the giant core ( which contains most of the energy ) has not been significantly affected by the distant influence of the companion .
once the companion has entered the envelope , which is still mostly intact , the orbital decay rate of the binary ( @xmath45 ) has almost reached its maximum value ( over a percent of the giant s radius per day ; see figures [ aa ] and [ adot ] ) .
over the course of the next 100 days the majority of the change in the core - companion separation occurs .
we define the end of this phase as the time at which the orbital decay timescale begins increasing again , after holding at a roughly constant value .
this occurs after approximately 400 days . in this simulation , about 90% of the remaining orbital angular momentum is transferred to the gas , mostly during the early part of the phase when the binary separation is decreasing most rapidly .
= 9 true cm = 9 true cm = 9 true cm = 9 true cm in the early portions of this stage , when the core - companion separation is decreasing most quickly ( on a timescale of about 100 days ) , the gas flowing past the companion is accelerated to supersonic velocity .
a bow shock is generated in this gas flow , directed more or less in the direction of travel of the companion , as is visible in the entropy contour plots in panel a of figure [ entfig ] .
later on when the point masses have spiralled closer , weak spiral shock waves can be seen in the density contours ( see figure [ rho6 ] ) . at first , only one shock is generated in front of the more rapidly moving companion . as a result , mass is pushed away from the giant in a very asymmetric pattern . even later
, shock waves are generated in front of both point masses .
near the point masses ( before the waves have gone through one - quarter of a turn ) , the shock waves are strongest , in the sense that more entropy is generated in the gas .
when the shock waves finally propagate into regions of lower density , the mach number of the gas increases , and they become more visible in the entropy plots ( as in panel b of figure [ entfig ] ) .
these shock waves are the primary means for transporting angular momentum away from the point masses during this phase .
= 18 true cm spiral shock waves continue to move outward from the point masses throughout this phase . because the period of the binary orbit is considerably smaller than the orbital timescale of the envelope gas , the spiral waves become tightly wound .
the angular momentum input from the shocks causes the mass distribution to elongate in the xz - plane , as seen in panel c of figure [ rho6xz ] .
the peak energy transfer occurs shortly after this point as the companion begins to dissipate energy in the high density gas at the core of the giant . at the same time , the angular momentum transfer from the point mass orbits to the gas is shutting off . the gas near the two cores is torqued up , but only a very small region ( @xmath46 cm in diameter ) following the companion s spiral wave reaches angular velocities near corotation . at the very end of the phase ,
the spiral shocks are very weak everywhere except in the vicinity of the point masses . outside this region ,
the gas motions are close to radial , which is the direction normal to the spiral waves .
a thick disk - like structure begins to take shape towards the end , and a low density region forms along the rotation axis ( see panel d of figure [ rho6xz ] ) .
the peak energy transfer from the orbits of the companion and the core to the gas occurs when the two particles are deep in the potential well of the system where more energy is available per unit change in separation .
very little of this energy has been converted into kinetic energy of the gas though . from figure [ mass1 ]
, we infer that most of the energy lost from the orbit is used in powering a relatively slow expansion of the mass closest to the core and companion . at the end of the phase there is little mass ( @xmath47 ) within a distance of several times the orbital separation from the point masses
more than half of the original content has been pushed outwards . on the other hand , the amount of mass within the original radius of the star
has decreased by about 30% .
this trend continues into the next phase of the evolution . through the midpoint of this phase ,
the specific entropy profile increases monotonically with cylindrical radius ( see figure [ sprof ] ) . by the end of the phase
, a small amount of mass ( @xmath48 ) near the core has developed a profile that decreases outward , which indicates convective instability .
this region is maintained during the remainder of the simulation , as gas motions apparently do not have time to redistribute the entropy .
( in the next phase of the evolution , we do see some evidence of slightly higher entropy blobs in the vicinity of the point masses .
this seems to indicate the presence of convection , although the motions of these blobs tend to keep them fairly close to the orbital plane . )
at this point , we also find the first indications of circulation perpendicular to the orbital plane with timescales on the order of a year .
this entropy profile around the point masses comes into being only when they take up close orbits around each other , and are able to repeatedly shock the gas . at about the same time
, the specific angular momentum of gas in the vicinity of the point masses also drops by a large amount ( see figure [ jprof ] ) , while much of the gas outside this region gains significantly .
this appears to be the result of the flattening of the giant .
= 9 true cm = 9 true cm = 9 true cm = 9 true cm at the end of this phase , the point masses are in close , relatively stable orbits about each other .
the center of mass of the orbiting point masses has received most of its `` kick '' from the ejected material also a speed amounting to just 3 km s@xmath49 .
the cores begin oscillating around the system s center of mass due to an asymmetry in the gas distribution , probably resulting from the spiral outflow pattern . the majority of the gas is still found within the original volume of the giant , but it continues to flow outwards as a result of the energy input from the core and companion .
the energy transfer rate ( shown in figure [ edot ] ) , which reached a peak value of @xmath50 , has declined by about half .
further expansion of the gas is facilitated through the use of a larger proportion of energy taken from the internal energy of the gas .
the removal of mass from the vicinity of the point masses was initially driven by spiral shocks , but at the end of this phase the reduced mass near the cores has considerably reduced the efficiency of angular momentum transport by them .
regardless , the gas is distributed primarily in the orbital plane of the binary as a result of the earlier action of these waves .
= 9 true cm = 9 true cm at the beginning of this phase , the cores have completed the majority of the orbital decay with the separation reaching @xmath51 cm .
the orbital decay timescale increases dramatically beyond this point .
this is primarily because the mass in the immediate vicinity of the two cores has declined to @xmath52 .
there was originally about @xmath53 of gas in this region of the red giant initially amounts this small are typical of the mass - radius profiles of red giants ( see , for example , figure 6 of yorke at al .
1995 , as well as figure [ mrd ] ) .
we note , however , most of the mass loss from the original volume of the giant ( which we will call the `` ejection '' ) does not occur until after the cores have established themselves in close orbits about each other .
= 9 true cm = 9 true cm the gas heated by close interaction with the two cores begins expanding away , and continues expanding through the remainder of the simulation , as shown in panels e and f of figure [ rho6 ] .
the rotational pattern associated with the disk - like structure seen in panel d is quickly overwhelmed by the outflow .
the expansion of the gas is driven by continued energy input from the point mass orbits , and is responsible for the movement of most of the remainder of the giant s mass out of the star s original volume .
this energy input also appears to be the cause of the circulations that develop perpendicular to the plane of the orbit within about @xmath54 cm of the point masses ( see especially panels e and f of figure [ rho6xz ] ) , confirming earlier work by taam & bodenheimer ( 1991 ) in two dimensions and terman & taam ( 1996 ) and rasio & livio ( 1996 ) in three dimensions .
the majority of the gas on the grid is still to be found within about one giant radius of the orbital plane . by 530 days ( panel f in the density plots )
, two thin cones having strong inflow are set up along the rotation axis .
these inflows complete the circulation pattern , partially replenishing the gas surrounding the point masses that is being driven out .
the strongest outflow is to be found close to the orbital plane , where gas can be accelerated to faster than the local sound speed . from figure [ mass1 ]
it is apparent that there is a net outflow , as can also be gathered by the fact that the strongest inflow involves relatively low density gas along the rotation axis . at later times ,
the outflow becomes more confined to the orbital plane ( @xmath55 cm ) . at the end of the simulation , over 90% of the gas
was pushed out of its original volume , and mass was still being removed from the volume at a rate of approximately 0.3 . about 45% of the gas had been pushed off the edge of the grid , although almost all of what remains is still bound ( although marginally ) to the system . at the end of the simulation ,
a significant amount of bound mass had been lost off the grid as a result of the expansion of the gas , making continued calculations unreliable . as was found at the end of the rapid infall phase
, the majority of the gas does not have angular velocity more than 10% of corotation . as seen in figure [ rhoomeg ] , a small amount of mass is maintained close to corotation very near the orbiting point masses , but the angular momentum transfer is negligible .
= 18 true cm in the end , the point masses have a separation of @xmath56 . at late times , the fractional decrease in the orbital separation of the cores should equal the fractional decrease in their gravitational potential energy since there is little envelope mass within the binary orbit .
we do find that the energy loss and orbital decay timescales go in lockstep . by the end of the simulation ,
the orbital decay timescale had increased to about 12 years .
the mass loss rate for the system ( computed from the amount of bound gas ) has remained roughly constant over most of the simulation , meaning the mass loss timescale has decreased to about 6 years by the end of the simulation .
a comparison of the timescales indicates that the entire common envelope should be ejected .
as can be seen from figure [ mass1 ] , the gas near the two cores is cleared out within about 100 days after the end of the rapid infall phase , and at the end of the simulation only about @xmath57 remains in the original volume of the giant . in all but the last of the following sequences ,
the companion was started on a circular orbit at a distance of @xmath58 cm away from the core of the giant . in all of the sequences but 2 and 5 , the envelope of the giant star was in synchronous rotation with the companion s orbit . in table 2 , we summarize various characteristics of the system at the end of the simulation : final separation of the point masses , their orbital period , the recoil velocity for the center of mass of the point masses , and the efficiency of envelope ejection @xmath59 .
[ we choose to use the definition @xmath60 ( tutukov & yungelson 1979 ) , where @xmath61 is the change in orbital energy of the point masses ( which is the difference in the sum of kinetic plus potential energies in the final state this was computed for the two point masses , but for the initial state , this was computed using the masses of the companion point mass and the giant ) , and @xmath62 is the binding energy of the mass ejected , as determined from the initial giant model . the mass ejected was taken to be the mass lost from the main grid plus unbound mass remaining on the grid minus bound mass lost from the main grid . ]
many details of the sequences are similar , so we will concentrate on the differences in the final orbital parameters and the mass - loss evolution . in this simulation ,
the initial conditions were chosen to be identical to those of the baseline simulation , except that the giant envelope was not rotating .
the expectation was that the larger relative velocity between the companion and the envelope of the giant would result in a greater deposition of the companion s orbital energy and angular momentum in the envelope .
a comparison of the mass tracers for the sequences ( figure [ mass1 ] ) indicates that the timescale for the companion to enter the envelope of the giant is actually slightly longer for the nonsynchronous case .
this is a result of less efficient angular momentum transfer for the nonsynchronous case in order to begin to enter the envelope , the companion s orbit must develop eccentricity , as can be seen in rasio & livio s ( 1996 ) calculation , and in ours .
in this simulation , the companion s gravitational torque acts for smaller amounts of time on any particular element of the surface of the giant s envelope .
mass is removed from the star s volume and from the grid at approximately the same rate for both simulation 1 and 2 .
we do not find any observable characteristics of the simulation that would allow us to determine whether the giant in such a binary was out of synchronism at the beginning of the interaction .
a simulation was run with a @xmath1 agb star to gauge the effects of different giant star mass ( with a more massive core and a more massive envelope ) on the evolution .
since the @xmath1 giant had nearly the same initial radius as the @xmath0 giant and the initial orbital separation was the same as in sequence 1 , the giant overfilled its roche lobe by a smaller margin .
this sequence was not followed as long as the baseline simulation ( 891 days , as opposed to 1027 days for sequence 1 ) , but at the end of the simulation the core - companion separation was found to be converging to approximately the same value .
although the binding energy of the envelope is larger , gas is lost from the original volume of the star and from the grid at nearly the same rate as in the baseline simulation , as shown in figure [ mass2 ] .
the implications are that the energy released from the orbit in these two simulations goes towards almost completely clearing out the region around the orbiting point masses , with the remainder ( probably the majority ) going toward expansion of the envelope in the orbital plane .
if , as appears to be the case , the orbital decay of the point masses is physically slowing ( as opposed to slowing due to numerical effects ) , we would expect that the envelope of the giant would not be completely ejected . at the end of this simulation
the orbital decay timescale is about 4 years , which is roughly half what it was at the comparable point in simulation 1 .
in addition , the orbital decay timescale is increasing at roughly half the rate it was in simulation 1 .
the mass - loss timescale at the end of the simulation is much longer ( @xmath63 years ) than the orbital decay timescale , further suggesting that the envelope of the giant might remain close enough to the point masses to affect their orbits further .
the inability of the companion to eject the more massive envelope is also reflected in the efficiency factor @xmath59 for this simulation we find that is only 24% , compared to @xmath64 40% for sequences 1 , 2 , and 4 . however , as discussed in
[ numerical ] , a different algorithm for the smoothing length in the gravitational potential leads to considerably more mass ejection and a higher efficiency factor , so these results for this simulation should be treated with care . in this simulation
, we again find that the companion s orbital decay has stopped at approximately the same distance from the giant core as it did in the baseline simulation .
the flat portion of the initial mass - radius profile of the @xmath1 giant had very nearly the same spatial extent as that of the @xmath0 model , which may explain the similarity if indeed the orbital decay decelerates when the two point masses begin orbiting within this region . at the end of the simulation ( which lasts 742 days )
, the orbital decay timescale has almost reached 6 years , indicating that the mass near the point masses has been efficiently removed .
the timescale is actually increasing faster than in simulation 1 . in examining the mass tracers , we find that the same _ fraction _ of the giant envelope is removed in half the time of the baseline simulation . at the end ,
the mass - loss timescale as judged from the bound mass is about 17 years and decreasing
. however , most of this gas is marginally bound , and it is sufficiently far away that it does not influence the point - mass orbits in the short term .
the mass in the original giant volume is decreasing considerably more quickly than in simulation 3 . as a test of the effects of the evolutionary state of the giant on the evolution
, we conducted this simulation in which the 5 giant was in a more evolved state .
the input giant model had a radius of @xmath65 cm , as opposed to @xmath66 cm for the previous runs with 5 giants .
the carbon - oxygen ( co ) core was 0.87 , and 0.57 for the previous runs .
( the point mass had a slightly lower mass than for the previous runs because the point mass includes all the mass of the star up to the size scale of about the smallest zone .
even though the co core is larger , the overall expansion of the giant , reduces the mass relegated to the point particle . ) in addition , the orbit of the companion was moved out to a distance of @xmath43 cm .
the simulation was run for 1574 days , and the final separation of the point masses in this simulation was approximately twice that of the previous simulations .
the mass tracers are shown in figure [ mass3 ] .
because this could have been related to our numerical algorithm , we explored this possibility , as described in the next section .
the result was that the smoothing length was found to affect the final separation , but not the efficiency of envelope ejection , which was found to remain roughly constant near 53% .
= 12 true cm the spiral - in phase lasts for the first 300 days in this simulation , while the rapid infall phase also lasts roughly twice as long as in the other simulations ( ending at about 800 days ) . at the end , the mass loss timescale has increased to 16 years , while the orbital decay timescale is at 7 years , and both are increasing .
the final outcome of the simulation wil probably be ejection of the envelope since the gas within the giant s volume is being pushed out quickly .
the later portions of the envelope ejection phase were not followed by rasio & livio ( 1996 ) because the local spatial resolution for their sph gas particles near the cores was comparable to the size of the binary orbit . in our simulations , limitations on spatial resolution
are present in the zone spacing on the grid the point masses occupy , and in the smoothing length applied to gravitational interactions between point masses and gas . with regards to the effects of the smoothing length in the gas - to - point - mass potentials ,
the main questions are how critical the formulation of the potential is , and how physically realistic it is .
the potential we have used crudely mimics the effects of constant density mass distributions in each gas zone , since the force due to a given zone is reduced when the point mass falls within the zone .
the point masses are not strictly points , although their physical sizes ( in the case of compact objects , including the cores of giant stars ) are more than an order of magnitude smaller than the smoothing length . however , relatively small amounts of matter accreted onto these stars will cause the cores to partially fill their respective roche or tidal lobes , creating configurations that are much larger ( see hjellming & taam 1991 ) and comparable to the zone size .
we expect that the smoothing length only becomes important towards the end of the simulations when the point mass orbits are small , meaning that the point masses orbit within and primarily interact with only a handful of zones . to test the effects of the smoothing length , we allowed it to decrease once the point mass separation reached a distance of three times the smoothing length of one point mass . to an extent this is physically realistic if the `` true '' size of the point masses are on the order of the tidal radius , since the tidal radii will decrease as the orbital separation decreases . in making this change for the configuration of sequence 5 , we find that the final orbital separation decreases from about 8.9 to 5.3
. we also ran the same simulations with a reduction in the number of subgrids from two to one to check the relative importance of the grid resolution itself . for a fixed potential smoothing length 1.5 times the smallest zone size ( twice the size of sequence 4 )
, we found the final separation was about 16.9 . when the smoothing length was fixed equal to the smallest zone size ,
the separation became about 14.9 . with the adjustable smoothing length algorithm ,
the final separation reached 8.1 . for simulations with additional subgrids ,
we expect the computed final separations would converge on a value . however at our present level of resolution ,
we are unable to specify exactly how close our two point masses approach each other in the end
. we should consider whether we can put realistic limits on the final separation .
it is safe to say that the separations derived from our fixed smoothing length simulations are good upper limits since higher resolution would tend to increase the amount of energy dissipation and angular momentum transfer .
we applied the newer smoothing length algorithm to simulations 1 , 3 , and 4 , and found that the final separation did decrease in all cases , but by no more than 25% .
the additional energy input into the gas leads to to the unbinding of considerably more mass and more mass loss off of the grid , while the amount of bound mass loss off the grid remains about the same . for simulation 3 , the additional decrease in orbital separation and the resultant mass loss might be enough to eject the envelope .
the fact that the orbital separations still end up being nearly the same with the new smoothing length algorithm may indicate that the grid resolution has become the most important factor .
we must conclude that it is unclear exactly how close the adjustable smoothing length algorithm brings us to the real physical separation .
there are a few final points we wish to make .
first , the first two stages of the common envelope evolution are unaffected by this question , and most of the details of the envelope ejection phase will not change .
for example , the efficiency of envelope ejection probably will not change much since the final separation is implicitly taken into account in the calculation .
in fact , for several test versions we ran of sequence 5 , the efficiency only varied by a few percent , despite differences in the final orbital separation and orbital energy of more than 50% .
second , in our simulations with fixed smoothing length , there is considerable eccentricity of the point mass orbits .
this shows up in the rapid variations of the point - mass kinetic and core - companion gravitational energies in the late stages of the simulations .
when we ran the same simulations with the smoothing length allowed to shrink with the orbit , the eccentricity is damped out .
we believe this damping will be observed in higher resolution computations .
the anonymous referee points out that our conclusions about the degree of corotation of the gas immediately surrounding the the two point masses may be affected by resolution effects .
rasio & livio ( 1996 ) found that reducing the resolution ( by reducing the number of particles ) suppressed corotation of the gas at the end of the simulation .
an examination of the present work and that of rasio & livio indicates that the two are not inconsistent with regards to the rotational state of gas farther from the cores . while the differences near the cores may be a function of numerical resolution , the results of the two studies indicate that the region of corotation is rather small in spatial extent ( of the order of the binary separation ) and negligible in mass , and hence dynamically unimportant .
there are a number of physical effects that have not been included in our models of the common envelope evolution , but which may affect either our results , or our inferences about the eventual outcome of the interaction . in the following , we briefly discuss several .
as discussed in the previous section , we have modeled the core of the red giant and the companion star as point masses . as a result
we neglect aerodynamic drag effects caused by the finite size of the stars and the gas they have accreted .
this would only tend to influence the results during phases in which the density of gas impinging on the stars significantly affects the momentum of the stars .
so , this probably does not affect the orbital evolution of the stars beyond the ends of our simulations , but could influence the rate at which the companion initially falls through the giant s envelope , or the orbital decay before too much of the envelope is ejected from the vicinity of stars .
even so , for objects traveling at supersonic speeds through a gas , aerodynamic drag is found to be small in most cases compared to the gravitational drag ( kley , shankar , & burkert 1995 ) , which we do model .
thus , we do not expect aerodynamic drag to play a large role .
it is also certain that radiation transfer becomes important at times beyond the end of the simulations .
as the gas density of the nebula drops , the decrease in optical depth allows the gas to cool , removing internal energy that could have been used to unbind it .
in addition , radiation from the white dwarf core of the red giant could help to power a wind that could assist in removing the remnants of the envelope .
if gas is able to cool and contract back towards the cores , it could result in additional reduction of the core - companion separation .
a rough calculation applying an analytic expression for bound - free opacities to the density and temperature distribution indicates that radiation diffusion timescales from the vicinity of the point masses are on the order of a thousand years at the end of the simulation .
it seems clear that gas must rarefy considerably before radiation losses become important .
in this paper the hydrodynamical evolution of a binary system consisting of an agb star and its main sequence companion has been followed from the initial to the late stages of the common envelope phase using high resolution numerical techniques . during the evolution ,
spiral density waves clearly show the coupling between the gas and the double core .
the shrinkage of the companion orbit is very rapid , lasting only about 0.7 year .
the orbital decay decelerates dramatically at an orbital separation of @xmath67 , although this may be due in part to the effects of finite spatial resolution .
the double core is briefly embedded in a slowly expanding differentially - rotating pressure - supported disk , which remains loosely bound to the two cores .
the matter within the original volume of the common envelope is spun up to only a fraction of corotation with the point masses , but energy input from the point mass orbits leads to mass removal at a rate of @xmath68 at the end of the simulation .
the evolution to the late stages was followed for about an order of magnitude longer in time than rasio & livio s ( 1996 ) sph calculation .
the results of our studies show that the late phases of the hydrodynamical evolution are not affected by the initial spiral - in of the companion into the giant star .
that is , it is found that the degree to which the components are out of synchronous rotation at the onset of the common envelope phase only affects the initial development .
a slightly less rapid infall of the companion into the red giant envelope results .
the final orbital separation of the remnant binary immediately following the ejection of the common envelope has not yet been reached in our simulations because of inadequate spatial resolution . for our first four simulations , a period of one day
is found to be the upper limit . with a more evolved giant
, the final period appears to increase .
the results of the numerical simulations suggest the successful ejection of the common envelope for a binary consisting of a @xmath69 giant with a @xmath70 core and a @xmath2 companion . on the other hand ,
the system may or may not merge for a higher mass giant of @xmath1 ( with a @xmath71 core ) with the same companion . the tendency toward merger for higher mass systems reflects the larger mass that must be ejected and the greater potential well from which the common envelope must escape . for more advanced stages in the evolution of the giant ,
a greater fraction of the mass is situated at larger distances from the degenerate core so that the energy requirements for ejection can be relaxed .
however , the flat mass - radius profile extends to greater distances as well , implying that the final orbital separation will also be increased , so that less orbital energy may actually be available for ejecting the gas .
thus , successful envelope ejection for more evolved configurations is not as favorable as might be expected . in all of our simulations
the mass ejection is concentrated toward the equatorial plane with about five times as much mass loss in the orbital plane as compared to the polar direction .
in addition , the results indicate that an expanding disk can form briefly about the remnant binary system
. these morphologies may be consistent with the equatorial - to - polar density contrast required by the interacting winds model for the shaping of planetary nebulae . specifically , the results described in frank et al .
( 1993 ) and mellema & frank ( 1995 ) indicate that variations of three to five are adequate in providing an envelope morphology which resembles those planetary nebula systems exhibiting a clear axisymmetric structure .
furthermore , the presence of a geometrically thick disk may facilitate the formation of jets in planetary nebulae ( soker & livio 1994 ) .
the results obtained from these numerical simulations can be used to place constraints on the progenitor systems of planetary nebulae with binary nuclei , provided that orbital angular momentum losses from the remnant binary are insignificant following the common envelope phase .
angular momentum losses associated with either magnetic braking or gravitational radiation are unimportant for the binary stars within planetary nebulae since the age of the nebula ( and hence the age of the post - common - envelope system ) is characteristically less than about @xmath72 years .
orbital evolution due to angular momentum transfer from the orbit to a circumbinary disk is also not expected to be significant since our results demonstrate that the majority of mass in the disk is situated at great distances ( many times the orbital separation ) from the binary system .
furthermore , the angular momentum loss from the orbit responsible for the spin - up and ejection of the remaining matter in the vicinity of the remnant binary ( @xmath73 ) can be estimated from the study of shu , lubow , & anderson ( 1979 ) , and it is expected to reduce the orbital separation only slightly ( by about 10% ) .
the observational study of binary stars within planetary nebulae reveals that many systems ( 6 out of the 12 known , although this is affected by selection effects ) have orbital periods less than 0.7 days ( see bond 1995 ) .
a total mass for the remnant binary of 1 @xmath10 implies that the orbital separation following the mass ejection phase of the common envelope is @xmath74 .
based on the present study and following the work of terman & taam ( 1996 ) on the estimated final orbital separations of post - common - envelope systems , it is found that white dwarf masses @xmath75 are indicated for systems with orbital periods less than 0.7 days .
( for systems with orbital periods @xmath76 days , no such constraint can be placed on the white dwarf masses ) .
it should be noted that white dwarf masses in the range from about 0.46 - 0.56 @xmath10 can not be formed in the common envelope scenario since such core masses develop during the core helium burning phase of the red giant when the red giant is insufficiently large in radius to initiate mass transfer ( see webbink 1976 ) .
the progenitor masses leading to systems with orbital periods @xmath77 days are less than @xmath0 as illustrated in figure [ taamfig ] .
based on the present work , the final orbital separations have been estimated by assuming that they are a factor of 6 less than the radial distance in the giant corresponding to the point where the logarithmic derivative of the pressure with respect to radius is a minimum ( this point marks the outer boundary of the flat portion of the flat mass - radius profile ; see terman & taam 1996 ) .
greater orbital separations correspond to progenitor stars in more advanced evolutionary phases with more massive cores .
the result that lower white dwarf masses result from the lower mass range of intermediate mass stars is consistent with the seminal work of paczynski ( 1970 ) .
the radii of these progenitor stars are @xmath78 cm corresponding to orbital periods of the progenitor systems at the onset of the common envelope phase of @xmath79 years . to further our understanding of the outcome of the common envelope phase , higher resolution and longer term evolutionary studies
will be necessary to follow the total ejection of the common envelope .
systematic studies will be required to determine the efficiency of mass ejection and to delineate the parameter space separating those systems which merge from those systems which survive the common envelope stage .
such studies will also illustrate the diversity of morphologies of ejected matter ( formation of disks and density contrasts between the polar and equatorial directions ) relevant to the shaping of planetary nebulae .
sequence & @xmath80 & @xmath81 ( days ) & @xmath82 ( km s@xmath49 ) & @xmath59 + 1 & 4.42 & 1.03 & 3 & 0.38 + 2 & 4.68 & 1.12 & 4 & 0.46 + 3 & 4.37 & 0.90 & 4 & 0.24 + 4 & 4.78 & 0.96 & 1 & 0.44 + 5 & 8.91 & 2.48 & 2 & 0.53 + | the evolution of binary systems consisting of an asymptotic giant branch star of mass equal to @xmath0 or @xmath1 , and a main sequence star of mass equal to @xmath2 or @xmath3 with orbital periods @xmath4 days has been followed from the onset through the late stages of the common envelope phase . using a nested grid technique , the three - dimensional hydrodynamical simulations of an asymptotic giant branch star with radii @xmath5 a.u .
indicate that a significant fraction of the envelope gas is unbound ( @xmath6% and 23% for binaries of @xmath0 and @xmath2 , and @xmath1 and @xmath7 respectively ) by the ends of the simulations , and that the efficiency of the mass ejection process @xmath8% . during an intermediate phase
, a differentially rotating structure resembling a thick disk surrounds the remnant binary briefly before energy input from the orbits of the companion and remnant core drive the mass away .
while the original volume of the giant is virtually evacuated in the late stages , most of the envelope gas remains marginally bound on the grid . at the ends of our simulations , when the orbital decay timescale exceeds about 5 years , the giant core and companion orbit one another with a period of @xmath5 day ( 2.4 days for a binary involving a more evolved giant ) , although this is an upper limit to the final orbital period . for a binary of @xmath1 and @xmath2 ,
the common envelope may not be completely ejected .
the results are not found to be sensitive to the degree to which the initial binary system departs from the synchronous state .
= -0.5 in epsf |
gan epitaxial layers grown on different substrates ( e.g. , al@xmath1o@xmath2 , sic , or si ) possess very large densities of threading dislocations which cross the layer along its normal , from the layer - substrate interface to the surface.@xcite the threading dislocation density depends only marginally on the substrate material ( and hence on the misfit between the substrate and the layer ) but rather on the growth technique and conditions . for ( 0001 ) oriented layers of wurtzite gan , the overwhelming majority of dislocations are of edge type with burgers vectors @xmath3 .
their density@xcite is typically @xmath4 @xmath5 .
the density of screw dislocations with burgers vector @xmath6 $ ] is one to two orders of magnitude lower than the density of edge dislocations . and
the diffracted wave @xmath7 make the same angle @xmath8 to the sample surface .
the scattering vector @xmath9 makes an angle @xmath10 to the sample surface . ]
the dislocation density can be measured directly by plan - view transmission electron microscopy ( tem ) .
the actual virtue of tem is not the accurate determination of the dislocation density , but rather the possibility to determine the type of the dislocation distribution ( e. g. , random vs. granular / columnar structure ) .
it is of limited statistical significance considering the small area covered by tem micrographs . alternatively
, x - ray diffraction can be used to detect the lattice distortions due to the presence of dislocations . in principle
, both the dislocation distribution and the actual dislocation density may be obtained from x - ray diffraction profiles .
the impact of screw and edge threading dislocations on the width of the x - ray reflections in the limiting cases of lattice planes parallel and perpendicular to the surface is commonly referred to as tilt and twist , respectively.@xcite this designation stems from the model of misoriented blocks@xcite which is _ not _ appropriate for strains caused by randomly distributed dislocations . in this case , the description in terms of mean - squared distortions@xcite is actually more adequate .
moreover , all previous experimental studies only use the full width at half - maximum ( fwhm ) of different reflections as a measure of the dislocation density .
the symmetric bragg reflections from gan layers are comparatively narrow , since they are not influenced by the presence of edge dislocations .
edge dislocations produce distortions within lattice planes parallel to the surface but do not disturb positions of these planes along the layer normal . the highest sensitivity to edge dislocations
is obtained by diffraction from lattice planes perpendicular to the surface .
this diffraction geometry requires grazing incidence illumination and is thus commonly performed at a synchrotron.@xcite an alternative geometry that can easily be realized in the laboratory is the skew geometry,@xcite which is a quasi - symmetric ( the incident and the diffracted waves have the same angles to the surface ) non - coplanar ( the surface normal does not lie in the plane defined by the incident and the diffracted waves ) geometry , as shown in fig.[skew ] . by measuring different reflections with increasing lattice plane inclination
, one can extrapolate to lattice planes perpendicular to the surface.@xcite a single reflection with a large inclination can be regarded as a figure of merit.@xcite a four - circle diffractometer@xcite is required for skew geometry measurements , since the sample has to be tilted with respect to its surface normal .
coplanar asymmetric reflections on a three - circle diffractometer are much less sensitive to edge dislocations since they only partially touch the lattice distortions parallel to the surface plane.@xcite the fwhm of the diffraction peak depends not only on the dislocation density , but also on the correlations between dislocations .
furthermore , it depends on the mutual orientations of the scattering vector , dislocation line direction , and the burgers vector direction , which have not been taken into consideration in previous diffraction studies of gan layers . in the present paper
, we analyze the entire lineshape of the diffraction profiles , and particularly their tails .
these tails are due to scattering in the close vicinity of the dislocation lines and are not influenced by the correlations between dislocations .
they follow universal power laws and can be used to determine the dislocation density .
we fit the entire diffraction peak profile to the numerical fourier transform of the pair correlation function and simultaneously obtain dislocation densities and the range of dislocation correlations .
x - ray diffraction is a well - established technique to analyze crystal lattice imperfections.@xcite the conventional and widely used methods ( also in gan studies@xcite ) are based on a comparison of the diffraction peak widths of different reflections.@xcite they aim to separate two contributions to the peak broadening , the finite size of the crystalline domains ( grains ) in the sample and the non - uniform strain within each domain owing to lattice defects .
the strain broadening of a diffraction peak is proportional to the length of the scattering vector , while the size effect does not depend on it .
thus , comparing the peak widths of reflections of successive orders in the so - called williamson - hall plot , one can separate both contributions .
the separation method assumes that all peaks are gaussian,@xcite or all lorentzian.@xcite this is certainly not true , and the methods based on the peak width give only rough , albeit instructive , estimates of the crystal perfection .
a recent development @xcite includes corrections for given orientations of the dislocation lines and burgers vectors with respect to the scattering vector .
the next milestone was the fourier analysis of the diffraction peak profile proposed by warren and averbach.@xcite the starting point of their analysis is the average over grain orientations in powder diffraction .
the corresponding integration of the scattered intensity in reciprocal space is equivalent to a one - dimensional cut of the correlation function in real space .
we note that the powder diffraction literature does not use the term correlation function but refers to the fourier coefficients of the intensity .
we use the solid - state physics terminology where this quantity is commonly called the pair correlation function . in crystallography , the same function is referred to as patterson function .
the ansatz of the warren - averbach analysis is the assumption that the correlation function is a product of two independent terms describing the size and the strain effects , respectively .
furthermore , it is assumed that the finite sizes of the grains give rise to a lorentzian peak ( exponential correlation function ) while non - uniform strain is described by a gaussian function . with these assumptions , the size and strain effects
are separated by differentiation of the fourier - transformed peak profile . in many cases ,
the latter step is not sufficiently accurate , since it has to rely on a few fourier coefficients obtained from noisy experimental data .
balzar@xcite suggested to avoid this difficulty by directly fitting the peak profile to a convolution of a lorentzian and a gaussian , which is the voigt function .
the experimental peak profiles are not always described by the voigt function and various other analytical functions were suggested on a purely phenomenological basis.@xcite the assumption that non - uniform strain gives rise to a gaussian correlation function looks plausible , taking into account the stochastic nature of this strain originating from randomly distributed lattice defects .
however , krivoglaz and ryboshapka@xcite have shown that this is not true for random dislocations , which are the most common and most important source of strain .
rather , they found that the slow ( @xmath11 ) decay of the strain with the distance @xmath12 from the dislocation line gives rise to a correlation function @xmath13 here @xmath14 is a dimensionless factor depending on the orientations of the dislocation line and burgers vector with respect to the @xmath15-direction ( an arbitrary direction along which the correlations are measured ) .
its dependence on the scattering vector @xmath9 and the burgers vector @xmath16 is given by @xmath17 .
the dislocation density @xmath18 is defined as a total length of the dislocation lines per unit volume . for straight dislocations , @xmath18 is equal to the number of dislocations crossing the unit area of a plane perpendicular to the dislocation lines .
@xmath19 is another dimensionless factor , for uncorrelated dislocations @xmath20 . for completely random and uncorrelated dislocations , @xmath21 is the crystal size , so that the diffraction peak width of an infinite crystal with random uncorrelated dislocations tends to infinity .
wilkens@xcite pointed out that this divergence has the same origin as the divergence of elastic energy of a crystal with dislocations : the elastic energy is proportional to @xmath22 , where @xmath23 is the shear modulus , @xmath24 is the length of the burgers vector , and @xmath25 is the lattice spacing .
he suggested that the system can drastically reduce elastic energy by a very minor rearrangement in the dislocation ensemble : the positions of the dislocations remain random but their burgers vectors are correlated , so that the total burgers vector in a region exceeding some characteristic scale is zero .
then , the dislocations screen each other and the elastic energy , as well as the diffraction peak width , remain finite .
wilkens introduced a `` restricted random distribution '' of dislocations by subdividing the crystal into cells , such that the total burgers vector in each cell is zero .
he found@xcite that the functional form of eq .
( [ eq1 ] ) does not change but @xmath21 should be understood as a finite size of the cells .
_ et al.__@xcite showed that the same result is valid for a broad class of correlated dislocation distributions with screening .
we note also the two - dimensional crystal as a limiting case of dislocation screening . here
the elastic energy of dislocations is to be compared with the entropic term @xmath26 , where @xmath27 is temperature and the entropy @xmath28 is the number of lattice sites where the dislocation can be placed .
both elastic energy and entropy terms are proportional to @xmath29 . as a result , above some temperature @xmath30 ,
dislocations are generated by unbinding of thermally excited dislocation pairs , giving rise to the dislocation mediated melting.@xcite the calculation of the correlation function in this highly correlated dislocation system@xcite shows that the logarithmic term in eq.([eq1 ] ) vanishes .
fourier transformation of the correlation function ( [ eq1 ] ) yields a gaussian shape only in the central part of the peak .
the range of the gaussian peak shape is given by a dimensionless factor @xmath31 and increases when @xmath32 is increased .
the intensity distribution notably deviates from the gaussian shape at the tails of the diffraction peak .
the tails of the diffraction peak due to dislocations follow a universal law @xmath33 , which can be obtained from fourier transformation of eq .
( [ eq1]).@xcite the @xmath0 law is due to the fact that at large @xmath34 , the scattering takes place in the strained regions close to dislocation lines , where the lattice is so strongly misoriented that the bragg law is locally satisfied for the wave vector @xmath34 .
calculation of the volume of these regions give the @xmath0 dependence.@xcite groma and co - workers@xcite developed methods for the peak profile analysis based on a calculation of the restricted moments of @xmath35 .
in particular , the second - order restricted moment @xmath36 is proportional to @xmath37 , which they used to determine the mean dislocation density .
higher - order moments describe fluctuations of the dislocation density .
gan epitaxial film comprise a well - defined system where threading dislocations are aligned perpendicular to the surface plane .
the film is oriented , contrary to a powder . however , the x - ray diffraction measurements performed with an open detector give rise to an average very similar to the powder average .
the intensity is integrated over directions of the outgoing beam , instead of the integration over directions of the diffraction vector .
the integrations in reciprocal space gives rise to cuts in real space , which are different for the two cases under consideration .
the coordinate @xmath15 in eq .
( [ eq1 ] ) runs along the direction of the outgoing wave in case of the oriented sample with open detector and in the direction of the diffraction vector for the case of powder diffraction .
this fact introduces a geometrical correction in the orientational factor @xmath38 .
the skew diffraction geometry used for the measurements gives rise to further corrections , which are calculated below .
our approach consists of a direct fit of the measured intensities by the numerical fourier transformation of the correlation function ( [ eq1 ] ) , thus avoiding any transformation of the experimental data .
we expect that such a calculation is less influenced by the noise in the experimental data and is more reliable .
as shown below , we find good agreement between measured and calculated peak profiles . from the fits ,
we reliably obtain both the dislocation densities and the correlation range in the restricted random dislocation distribution .
the intensity of x - ray scattering from a crystal disturbed by strain fields of lattice defects can be represented as a fourier transform @xmath39 of the pair correlation function @xmath40\}\right\rangle . \label{eq3}\ ] ] here @xmath41 is the scattering vector ( @xmath42 and @xmath43 are the wave vectors of the incident and scattered waves , respectively ) and @xmath44 is a small deviation of @xmath9 from the nearest reciprocal lattice vector @xmath45 , so that @xmath46 .
@xmath47 is the sum of displacements produced by all defects of the crystal in a given point @xmath48 and @xmath49 denotes the average over the statistics of the defect distribution .
equation ( [ eq3 ] ) implies an infinite and statistically uniform sample , so that the choice of origin is arbitrary . when finite size effects are essential , as for example for misfit dislocations in epitaxial layers,@xcite the correlation function @xmath50 depends on the difference of displacements @xmath51 and eq .
( [ eq2 ] ) contains the exponent @xmath52 $ ] .
we restrict ourselves to parallel straight dislocations and take the direction of the dislocation lines as @xmath53 axis .
then , @xmath54 is a two - dimensional vector in the plane perpendicular to the dislocation lines and eq .
( [ eq2 ] ) can be written as @xmath55 where the delta - function @xmath56 is due to the translational invariance in the direction of the dislocation lines . in the experiments described in the subsequent sections ,
the x - ray scattering measurements from oriented samples are performed with a wide open detector .
the intensity ( [ eq2 ] ) is then to be integrated over all directions of the scattered wave @xmath43 .
the result of this integration is very similar to the powder average .
the actual part of the sphere @xmath57 ( where @xmath58 is the wave vector ) can be replaced by the plane perpendicular to the direction of @xmath43 .
integration of the intensity ( [ eq2a ] ) over this plane gives rise to a one - dimensional integral @xmath59 where @xmath60 is the angle between the @xmath61 plane and @xmath62 ( see fig . [ skew ] ) .
it is given by @xmath63 , where @xmath64 is the angle between the @xmath61 plane and the scattering vector @xmath9 and @xmath65 is the bragg angle .
the @xmath15 axis is chosen along the projection of @xmath66 on the plane perpendicular to the dislocation lines .
the corresponding expression for the case of powder diffraction differs only by the direction of @xmath15 : it runs along @xmath43 for oriented films and along @xmath9 for powder diffraction .
the wave vector @xmath34 in eq .
( [ eq7 ] ) is the projection of @xmath67 on the direction of @xmath43 , so that @xmath68 , where @xmath69 is the angular deviation from the peak center .
krivoglaz@xcite performed the average over an uncorrelated defect distribution and showed that the correlation function can be represented as @xmath70}\right ] d\mathbf{r}\right\ } .
\label{eq4}\ ] ] here @xmath71 is the displacement field produced at a point @xmath48 by a single defect located at the origin and @xmath18 is the defect density . for straight dislocations , the integration in eq.([eq4 ] ) is performed over the @xmath72 plane perpendicular to the dislocation lines and @xmath18 is the dislocation density per unit area .
thus , according to eq .
( [ eq4 ] ) , in order to obtain positional correlations between two points separated by a distance @xmath48 , one has to place a dislocation in an arbitrary position @xmath73 and perform the integration over all @xmath73 . as a result of the slow decay of the dislocation strain , the main contribution to the integral is due to remote dislocations , @xmath74 , so that the difference of displacements can be expanded as taylor series , @xmath75\approx ( \mathbf{r}\cdot \nabla ) [ \mathbf{q}\cdot \mathbf{u}(\mathbf{r } ) ] = r_{i}q_{j}\partial u_{j}/\partial x_{i}$ ] .
the distortion field of a dislocation has a universal @xmath12-dependence , @xmath76 , where @xmath24 is the length of the burgers vector and @xmath77 is a dimensionless factor of the order of unity which depends only on the azimuth @xmath78 .
if the densities of dislocations with opposite signs of the burgers vectors are equal ( there is no plastic bend ) , the imaginary part of the correlation function ( [ eq4 ] ) is zero and its real part , after expanding the exponent in curly brackets , becomes@xcite @xmath79 where @xmath80 @xmath38 and @xmath81 are dimensionless factors of the order of unity . here
@xmath82 and @xmath83 are unit vectors in the directions of @xmath48 and @xmath9 , respectively .
the integral in eq .
( [ eq5 ] ) is taken from @xmath84 ( where @xmath19 is a dimensionless factor ) to a limiting size @xmath21 , which , for completely uncorrelated dislocations , is equal to the crystal size .
then , the integral is equal to @xmath85 , and we arrive at eq .
( [ eq1 ] ) .
when the dislocations are correlated , so that the total burgers vector averaged over a certain characteristic scale is zero , the functional form of the correlation function does not change but @xmath21 has the meaning of that scale.@xcite the integration range in eq .
( [ eq7 ] ) is limited by distances smaller than @xmath21 . a finite integration limit introduced in eq .
( [ eq7 ] ) leads , in a numerical evaluation of the integral , to unphysical oscillations in @xmath86 commonly appearing in fourier integrals taken over a rigidly limited interval .
we found that an appropriate smoothing is obtained by substituting @xmath87 in eq .
( [ eq1 ] ) with @xmath88 $ ] and extending the integration range to infinity .
the expressions for @xmath89 are bulky and depend on the type of correlations in dislocation positions.@xcite we restrict ourselves to the first approximation that does not depend on the type of correlations , @xmath90 . finally ,
combining the equations above , the diffracted intensity can be represented as @xmath91 we proceed here to the angular deviation from the peak maximum @xmath69 and introduce the peak height @xmath92 and the background intensity @xmath93 to provide the exact formula that is used for the fits of the experimental peaks presented below .
parameters @xmath94 ( describing the dislocation density ) and @xmath95 ( describing the dislocation correlation range ) are given by @xmath96 here @xmath97 and @xmath98 are dimensionless quantities given by the diffraction geometry , @xmath99 in eq .
( [ eq11 ] ) , the expression for @xmath98 is written for edge dislocations , taking into account the approximation @xmath100 .
for screw dislocations , @xmath101 in this expression should be replaced by @xmath102 .
the length of the burgers vector @xmath24 in eq .
( [ eq10 ] ) is that of the relevant burgers vector for either edge or screw dislocations .
the dimensionless product @xmath103 is the mean number of dislocations crossing each @xmath104 cell in the plane perpendicular to the dislocation lines .
equation ( [ eq9 ] ) with four parameters , @xmath105 is used in sec .
[ sec : results ] below to fit the peak profiles and obtain the dislocation density @xmath18 and the length @xmath21 . ) for different values of the parameter @xmath106 .
all curves merge at a common @xmath0 asymptotic .
a gaussian peak profile is shown by the thin line for comparison . ]
the behavior of the integral ( [ eq9 ] ) is illustrated in fig .
[ asymptotics ] where the function @xmath107\}\cos ( qx)dx \label{eq12}\ ] ] is numerically calculated for different values of the parameter @xmath106 .
the curves merge at a common @xmath0 asymptotic that does not depend on @xmath106.@xcite then , eq.([eq9 ] ) has an asympotic behavior ( for @xmath69 large in comparison with the peak width ) @xmath108 we note that @xmath109 is the integrated intensity of the peak .
figure [ asymptotics ] also shows that the asymptotics ( [ eq13 ] ) is reached quite close to the peak center for @xmath110 , while for @xmath111 the central part of the peak is gaussian and the angular range where the gaussian approximation is valid increases with increasing @xmath106 .
the fwhm of the calculated peaks increases with increasing @xmath106 and can be approximated by @xmath112 groma@xcite suggested to use the second restricted moment of the intensity distribution @xmath113d\varpi \label{eq14}\ ] ] to obtain the dislocation density from the asymptotic behavior ( [ eq13 ] ) .
note that the integral ( [ eq14 ] ) diverges , when taken in infinite limits .
one finds , by substituting ( [ eq13 ] ) into ( [ eq14 ] ) , that @xmath114 it remains to calculate the orientational factor @xmath38 [ eq . ( [ eq6 ] ) ] . in the case of powder
diffraction,@xcite the vector @xmath48 is directed along the projection @xmath115 of the scattering vector @xmath9 on the @xmath61 plane . in our case ,
the vector @xmath48 is directed along the projection of @xmath43 on the @xmath61 plane and makes an angle @xmath116 with the vector @xmath115 , see fig .
this angle is given by @xmath117 .
when evaluating the integral ( [ eq6 ] ) for edge dislocations , we average @xmath81 over possible orientations of the burgers vector in a hexagonal lattice ( the dislocation lines are taken along the sixfold axis ) and obtain @xmath118 where @xmath119 is the poisson ratio .
@xmath120 depends only weakly on @xmath116 .
taking @xmath121 , one can approximate @xmath122 in the whole range of reasonable angles @xmath116 .
the calculation for screw dislocations gives @xmath123 two limiting cases are of interest : a symmetric bragg reflection to study screw dislocations and a grazing incidence / grazing exit reflection as an extreme case of a highly asymmetric skew geometry .
for a symmetric bragg reflection , @xmath124 and @xmath125 .
then , we obtain @xmath126 and , for screw dislocations , @xmath127 .
the grazing incidence / grazing exit geometry is the limit @xmath128 , so that @xmath129 and , for edge dislocations , @xmath130 .
the finite thickness @xmath27 of the epitaxial layer ( with threading dislocations perpendicular to the layer plane ) can be taken into account by making the warren - averbach ansatz@xcite @xmath131 where @xmath132 is the correlation function considered above .
the correlation function describing finite size effects can be written as @xmath133 .
its fourier transformation is a lorentzian that is expected from the finite slit function @xmath134^{2}$ ] after averaging over thickness variations .
we note that , in asymmetric reflections , eq .
( [ eq7 ] ) gives rise to the effective thickness @xmath135 along the direction of the diffracted wave .
the finite size correction ( [ eq17 ] ) does not complicate the calculation of the intensity by numerical integration of eq.([eq7 ] ) .
if the resolution of the experiment can not be neglected in comparison with the peak width , the correlation function ( [ eq17 ] ) is to be multiplied with the real - space resolution function @xmath136 , which also does not lead to additional complications of the numerical integration .
the gan layers investigated here were grown on 6h - sic(0001 ) by plasma - assisted molecular beam epitaxy ( pambe ) .
the two pambe systems employed are equipped with a solid - source effusion cell for ga and a radio - frequency nitrogen plasma source for producing active n. both systems have a base pressure of 5@xmath13710@xmath138 torr .
we use 6n n@xmath1 gas as a precursor which is further purified to 5 ppb by a getter filter .
h@xmath139-etched 6h - sic(0001 ) wafers produced by creewere used as substrates.@xcite an _ in situ _ ga flash - off procedure was performed in order to remove residual suboxides from the sic substrate surface prior to growth.@xcite the temperatures were calibrated by visual observation of the melting point of al ( 660@xmath140c ) attached to the substrate .
scan with analyzer ( gray line ) recorded in skew geometry across the 10.5 reflection of sample 1 . the symmetric 004 reflection from a perfect si crystal ( broken line ) is shown as a measure of the resolution .
the intensities from the gan layer are presented in counts per second , the si(004 ) signal is scaled appropriately .
the right panel shows the same peaks in log - log scale . ]
sample 1 was grown under ga - stable conditions with a substrate temperature of 740@xmath141c .
the growth rate employed was 140 nm / h , and the film was grown to a final thickness of 340 nm . under these growth conditions , we observed an entirely streaky , ( 1@xmath1421 ) rheed pattern throughout growth except for the initial nucleation stage .
the surface morphology of the film as observed by atomic force microscopy ( afm ) exhibits clearly resolved monatomic steps .
the root - mean - square roughness amounts to 0.3 nm over an area of @xmath143 @xmath144m@xmath145 .
samples grown under these conditions typically exhibit a narrow symmetric reflection , suggesting a low density of screw dislocations .
in contrast , sample 2 was grown under near - stoichiometric conditions with a substrate temperature of 780@xmath141c .
the growth rate employed was 445 nm / h and the total thickness of the film amounts to 1660 nm .
after the initial 100 nm of gan growth , a 10 nm thick aln film was deposited prior to overgrowth by gan .
the rheed pattern exhibited a superposition of streaks and @xmath146-shaped chevrons indicative of facetting .
indeed , the afm micrographs of this sample showed the characteristic plateau - valley morphology of gan films grown with insufficient ga flux .
compared to films grown under the same conditions as sample 1 , films grown under these conditions exhibit narrower asymmetric reflections , indicating a reduction of the edge dislocation density .
x - ray measurements were performed with a philips xpert pro four - circle triple - axis diffractometer equipped with a cu@xmath147 source in the focus of a multilayer x - ray mirror and a ge(022 ) hybrid monochromator .
the detector was kept wide open and at fixed position @xmath148 , leading to an angular acceptance of @xmath149 .
all asymmetric rocking curves were recorded in skew geometry.@xcite we denote the gan reflections in the form _
hk.l _ which is equivalent to the four - index notation _
hkil _ for hexagonal crystals with @xmath150 .
) . the dotted lines for the 10.1 profile show a gaussian profile .
] figure [ omegascan ] compares a double - crystal rocking curve measured with open detector ( acceptance 1@xmath140 ) and a triple - crystal @xmath151 scan measured with a three - bounce ge(022 ) analyzer across the 10.5 reflection for the same gan film ( sample 1 ) .
the comparison with the 004 rocking curve from a perfect si crystal shows that the gan peak broadening due to instrumental resolution can be neglected .
the log - log plots ( right column ) show that the tails of the intensity distributions follow the asymptotic laws expected for scattering from dislocations , @xmath152 for measurements with the open detector and @xmath153 in the case of a collimated diffracted beam .
thus , the comparison of these two profiles shows that they contain the same information about lattice distortions in the film .
the rocking curve measurements with open detector have , however , both experimental and theoretical advantages .
the experimental advantage is a two orders of magnitude higher intensity ( the intensities are plotted in fig .
[ omegascan ] in counts per second ) .
the analysis of the rocking curve is also more simple since the intensity distribution is described by the one - dimensional integral ( [ eq7 ] ) , while the analysis of the scans with the analyzer requires the two - dimensional integration ( [ eq2a ] ) of the correlation function .
therefore , we restrict ourselves to the analysis of double - crystal rocking curves .
and @xmath95 obtained from the fits of the experimental profiles by eq .
( [ eq9 ] ) .
samples 1 and 2 are denoted by full and open symbols , respectively . ]
figure [ asymm_peaks ] presents skew - geometry rocking curves for various reflections from sample 1 .
several conclusions can be drawn just from a direct inspection of the peak profiles .
first , the lineshapes are far from being gaussian at the tails of the intensity distribution : see the comparison of the 10.1 profile with a gaussian fit [ dotted lines in figs .
[ asymm_peaks ] ( a , b ) ] .
when the dynamic range is larger than two orders of magnitude , it is evident that the tails of the asymmetric profiles rather follow the @xmath0 law .
for relatively weak reflections ( e. g. , 10.4 or 20.4 ) , the @xmath154 exponent is not reached .
thus , the profiles obey the behavior typical for a random dislocation distribution . in the following ,
we analyze them quantitatively to obtain the characteristics of the dislocation ensemble . the solid lines in figs .
[ asymm_peaks ] ( a ) and ( b ) are the fits of the measured profiles by eq .
( [ eq9 ] ) .
one can see that the peak profiles are adequately described . in fig .
[ parametersab ] , we plot the fit parameters @xmath94 and @xmath95 as functions of @xmath97 and @xmath98 , respectively , since according to eq .
( [ eq10 ] ) both dependencies are expected to be linear .
figure [ parametersab ] ( a ) can be considered as a refined version of the williamson - hall plot .
the linear fit of the data in fig.[parametersab ] for sample 1 crosses the axis of the ordinates at a small but non - zero value of @xmath94 , which indicates that , in addition to threading dislocations , there is an additional source of peak broadening , namely , size broadening .
this effect is much smaller for sample 2 . from the slopes of the straight lines in fig .
[ parametersab ] ( a ) , we obtain @xmath155 and @xmath156 for samples 1 and 2 , respectively , where @xmath157 nm is the length of the burgers vector of edge dislocations .
this result yields a density of edge threading dislocations @xmath158 @xmath5 for sample 1 and @xmath159 @xmath5 for sample 2 .
the mean distances between edge dislocations are @xmath160 nm ( sample 1 ) and 53 nm ( sample 2 ) . from the slopes of the lines in fig.[parametersab ] ( b ) , we obtain @xmath161 and @xmath162 , which give the characteristic lengths of the dislocation correlations @xmath163 nm and @xmath164 nm , respectively .
the dimensionless parameter characterizing the dislocation correlations@xcite possesses the values @xmath165 for sample 1 and 2.0 for sample 2 .
we also fitted the profiles with eqs .
( [ eq13 ] ) and ( [ eq15 ] ) , which require less computational efforts .
these fits give slightly larger values for parameter @xmath94 ( dislocation density ) and are not able to give parameter @xmath95 for dislocation correlations .
for the two gan films .
the symbols are experimental data and the lines are fits to them . ] the symmetric x - ray diffraction profiles shown in fig .
[ symm_peaks ] are narrow compared to the asymmetric reflections . furthermore , the profiles of samples 1 and 2 are qualitatively different . in sample 1 ,
the intensity distributions obey a @xmath166 law in the intermediate range of angular deviations ( and intensities ) that is followed by an even lower exponent for large deviations ( and very low intensities ) . in sample 2 ,
the intensity distributions are close to the @xmath0 law .
edge threading dislocations with dislocation lines normal to the surface and burgers vectors in the surface plane , which are the main source of the peak broadening in asymmetric reflections , do not distort the planes parallel to the surface .
they do thus not contribute to diffraction in the symmetric reflections .
the @xmath166 law for the 00.2 and 00.4 peaks of sample 1 points to the finite thickness of the epitaxial layer as the main source of the broadening .
the @xmath167 law at the far tails of the peaks are due to another source , possibly thermal diffuse scattering .
we did not investigate this part of the symmetric reflections since we suppose that it is not related to threading dislocations which are the topic of the present study .
the profiles of sample 2 , which is three times thicker than sample 1 , obey the @xmath0 law indicating that the broadening is primarily by dislocations .
the solid lines in fig .
[ symm_peaks ] are the fits of the experimental curves to eq .
( [ eq17 ] ) , where the different thicknesses of the layers are explicitly taken into account .
we indeed obtain the @xmath166 intensity decay on the tails of the peak for sample 1 and the @xmath0 decay for sample 2 . from the fit parameters , we obtain @xmath168 and @xmath169 for samples 1 and 2 , respectively , where @xmath170 nm is the length of the burgers vector of screw dislocations .
this results yields the screw threading dislocations densities of @xmath171 @xmath5 for sample 1 and @xmath172 @xmath5 for sample 2 . from the values of the parameter @xmath95 we obtain @xmath173 and @xmath174 , which result in characteristic lengths of the dislocation correlations @xmath175 nm and @xmath176 nm for sample 1 and 2 , respectively .
the parameter @xmath32 is close to 1 for both samples .
note , however , that screw dislocations are not the only source of the peak broadening in symmetric reflections ( see discussion below in sec .
[ sec : discuss ] ) .
tem is the method of choice to directly determine the character of dislocations and their distribution in thin films .
the @xmath177 criterion in tem is generally applied for the two - beam condition to evaluate the alignment of the strain field of the dislocation with burgers vector @xmath16 with respect to the diffraction vector @xmath178 producing the image contrast .
this criterion is strictly correct for screw dislocations and for edge dislocations only if their line direction @xmath179 and burgers vector are in the imaging plane , i. e. , if @xmath180 is considered . in the present case of gan(0001 ) films it is well established that three types of threading dislocations exist , having burgers vectors @xmath181 , @xmath182 and @xmath183 representing edge , screw and mixed dislocations , respectively , under the assumption that the dislocations lines lie parallel to the @xmath184-axis .
mixed dislocations are not observed in our samples . in order to identify the burgers vector ,
two images have to be recorded with @xmath178 parallel and perpendicular to the @xmath184-axis .
the screw dislocations are thus imaged if @xmath178=[0002 ] , an example of which is shown in fig .
[ tem](a ) for sample 2 . from this image , we can directly measure the dislocation density if we know the tem specimen thickness that is determined by tilting the interface from the end - on to a well - defined inclined position .
edge dislocations appear in the cross - sectional images if @xmath178 is perpendicular to the @xmath184-axis , e. g. , if @xmath185 $ ] , and we are then able to measure their density in the same way .
furthermore , plan - view tem imaging is applied to complement the measurement of the dislocation density .
figure [ tem](b ) shows a plan - view tem image of sample 2 , where the specimen is tilted a few degrees off the [ 0001 ] zone axis to obtain two - beam conditions with @xmath185 $ ] in order to bring the edge dislocations in contrast .
the edge threading dislocations are extended along the film normal and homogeneously distributed .
it is remarkable that screw dislocations can be identified as well ( marked by arrows ) , although the contrast should vanish because of @xmath186 .
however , this specific contrast is produced due to strain relaxation of the screw dislocation at the free surfaces varying locally the lattice plane distortions created by the strain field .
table [ table ] compares the densities of edge and screw dislocations obtained from the x - ray diffraction and tem measurements , revealing that the density of edge dislocations is in fact slightly higher in sample 1 compared to sample 2 .
the screw dislocation density in sample 1 is too low for reliable tem determination .
some of the screw dislocations are seen as hexagonal pits in afm micrographs .
a lower limit of 5@xmath14210@xmath187 @xmath5 is thus obtained for the screw dislocation density in sample 1 .
.edge and screw dislocation densities determined by x - ray diffraction and tem .
[ cols="<,^ , < , < , < , < " , ]
the x - ray diffraction studies of gan have used , up to now , only the widths of diffraction peaks .
two quantities can be obtained from a series of the reflections.@xcite one is the fwhm of the symmetric reflection @xmath188 which is influenced by screw dislocations but insensitive to edge dislocations .
the other quantity is obtained by extrapolation of the fwhm of skew reflections to the limiting case of grazing incidence / grazing exit diffraction , with both incident and diffracted beams lying in the surface plane . this quantity ( @xmath189 )
is sensitive to edge but insensitive to screw dislocations .
figure [ twist ] presents the widths of the diffraction profiles of samples 1 and 2 as a function of the inclination angle @xmath10 .
the lines are fits by the model described in ref . to obtain @xmath188 and @xmath189 .
while these quantities are determined reliably and accurately , the question arises how they are related to the actual dislocation densities .
most commonly,@xcite the formulas initially proposed in refs . and for a mosaic crystal
are used for this purpose : @xmath190 with the coefficient @xmath191 in this context , the quantities @xmath188 and @xmath192 are often referred to as tilt and twist , respectively .
equations ( [ eq18 ] ) are valid for mosaic crystals , where the terms tilt and twist describe different modes of relative block misorientation .
these terms loose their meaning when applied to a layer with randomly distributed dislocations , since the broadening is then determined by a complicated combination of misorientation of the lattice planes and change of interplanar distances ( strain ) .
hence , equations ( [ eq18 ] ) use an inappropriate approach and are not strictly valid in the case of a random dislocation distribution .
it remains correct that the dislocation density is proportional to the square of the peak width and inversely proportional to the square of the relevant burgers vector , but the coefficients need to be reconsidered .
the relation between the dislocation density and the fwhm @xmath193 of the intensity distribution ( [ eq9 ] ) can be written , by making use of eqs .
( [ eq10 ] ) and ( [ eq13a ] ) , in a form similar to eq .
( [ eq18 ] ) : @xmath194^{2}fb^{2}}. \label{eq19}\ ] ] the dimensionless parameter @xmath195 that characterizes dislocation correlations is slightly larger than 1 for the gan films studied here , which is an indication of a strong screening of the long - range strain fields of dislocations by the neighboring dislocations .
one has @xmath196 for uncorrelated dislocations . in this latter case , the logarithmic term in eq .
( [ eq19 ] ) is the one obtained by krivoglaz.@xcite we can now simplify eq .
( [ eq19 ] ) for the limiting cases of the large - inclination skew diffraction and for symmetric bragg diffraction by using the expressions for the parameters @xmath97 and @xmath98 obtained in sec.[sec : theory ] for these cases : @xmath197 the term @xmath198 describing the range of dislocation correlations can not be neglected even when @xmath32 varies between 1 and 2 , as it happens to be the case for the samples studied in the present work .
compared to eqs.([eq18 ] ) , eqs .
( [ eq20 ] ) result in a four times higher edge dislocation density and an order of magnitude higher screw dislocation density and give dislocation densities that are in agreement with the tem results , see table [ table ] .
there are at least three types of dislocations that may contribute to the broadening of symmetric reflections .
first , screw threading dislocations distort the lattice planes parallel to the surface .
secondly , edge threading dislocations can contribute to these reflections if the dislocation lines deviate from the layer normal .
even if such deviations are small , the effect can not be neglected since the density of edge threading dislocations is much larger than the density of screw threading dislocations .
third , misfit dislocations at the layer - substrate interface contribute to broadening of the symmetric reflections .
although misfit dislocation lines are far from the surface , their non - uniform distortions penetrate through the whole epitaxial layer.@xcite finally , the stress relaxation at the free surface gives rise to additional strains around the outcrops of the edge threading dislocations , contributing to an additional broadening to the symmetric reflections .
thus , eq . ( [ eq9 ] ) and , particularly , eqs .
( [ eq20 ] ) provide an upper estimate of the screw threading dislocation density since they are assumed to be the only source of broadening for symmetric reflections .
the width of either symmetric or asymmetric reflection can be used as a figure - of - merit for the dislocation density only if the dislocation distribution is the same in all the samples to be compared . even for films having a spatially random distribution of dislocations ,
the width of a given reflection depends not only on the dislocation density , but also on the range of correlations in the restricted random dislocation distribution .
the lineshape analysis of the diffraction profile as presented in this work returns the width as well as the correlation range , and is thus a far more reliable approach for estimating the dislocation density than a simple consideration of the width alone . the lineshape analysis has shown that the x - ray diffraction profiles of the gan films under investigation are gaussian only in the central part of the peak .
the tails of the peak follow the power laws characteristic to x - ray diffraction of crystals with randomly distributed dislocations .
the double - crystal rocking curves measured with a wide open detector follow a @xmath0 behavior , while the triple - crystal rocking curves with an analyzer crystal obey a @xmath199 behavior .
the study of the double - crystal rocking curves is more simple both experimentally and theoretically , since the diffracted intensity is larger and the peak profile is described by a one - dimensional fourier transform of the pair correlation function ( [ eq9 ] ) .
the @xmath0 tails of the diffraction profiles are insensitive to correlations between dislocations and allow a more reliable determination of the dislocation densities .
the entire diffraction profiles are adequately fitted by eq .
( [ eq9 ] ) .
the fits provide two parameters characterizing the dislocation ensemble , the mean dislocation density @xmath18 and the screening range @xmath21 .
the latter quantity corresponds to a mean size of the cells with the total burgers vector equal to zero .
we find that , for edge threading dislocations in gan layers , @xmath21 is only slightly larger than the mean distance between dislocations @xmath200 . | we analyze the lineshape of x - ray diffraction profiles of gan epitaxial layers with large densities of randomly distributed threading dislocations .
the peaks are gaussian only in the central , most intense part of the peak , while the tails obey a power law .
the @xmath0 decay typical for random dislocations is observed in double - crystal rocking curves .
the entire profile is well fitted by a restricted random dislocation distribution .
the densities of both edge and screw threading dislocations and the ranges of dislocation correlations are obtained . |
it has long been known that t tauri stars ( ttss ) emit infrared ( ir ) radiation in excess of their stellar photosphere ( e.g. , * ? ? ? * ) .
@xcite speculated that silicate dust in orbit around these stars was responsible for this excess emission . with observations from the _ infrared astronomical satellite _ ( iras )
, it was shown that the 12100@xmath0 ir excess emission from these young stars could arise from dusty accretion disks @xcite .
many different models for this disk emission have been proposed .
both @xcite and @xcite construct disk models including both accretion and reprocessing of stellar radiation . in order to explain how disk reprocessing can be responsible for the ir excesses of most ttss , (
* henceforth kh87 ) proposed that disks around ttss are flared , in that the scale height of the disk increases more than linearly with the distance from the central star . a flared disk intercepts a larger solid angle of radiation emitted from the star than a flat or nonflared disk , leading to more reprocessing of starlight .
kh87 suggest that the surface of such a flared disk would become hotter than the midplane due to radiative transfer effects .
the disk material is optically thick to @xmath3@xmath41@xmath0 stellar radiation , so the starlight is absorbed in the highest layers of the disk . at @xmath3@xmath410@xmath0 , characteristic of the reprocessed radiation from the top disk layers at r@xmath41au in the disk ,
the disk has less optical depth , and the reprocessed radiation diffuses into the interior parts of the disk and heats those regions .
for the small accretion rates typical of tts , the disk atmosphere heats to a higher temperature than the layers in the disk underneath , and the vertical temperature inversion produces silicate features in emission @xcite .
@xcite summarizes how , based on the ir spectroscopic observations of @xcite , it came to be established that the 10@xmath0 emission ( or absorption ) feature , the broad emission ( or absorption ) feature from 8 to 13@xmath0 seen in a number of astronomical objects , is due to the si o stretching modes in silicate grains .
spectrophotometric observations by @xcite and @xcite at wavelengths longer than 16@xmath0 provided further support for the silicate hypothesis .
they found an 18.5@xmath0 peak in the trapezium emission and an 18.5@xmath0 maximum in the absorption from the bn - kl source ; the pairing of the 18.5@xmath0 feature ( the broad emission or absorption feature from 16 to 23@xmath0 ) with the 10@xmath0 features in the trapezium and the bn - kl source confirmed the silicate hypothesis .
there have been many studies of the silicate features of ttss , both ground - based @xcite and space - based @xcite . however , ground - based spectroscopic observations are limited by the earth s atmosphere at mid - ir wavelengths in both wavelength coverage and sensitivity . space - based missions , such as the _ infrared space observatory _ ( iso ; * ? ? ?
* ) do not suffer these limitations .
the _ spitzer space telescope _
@xcite offers greater sensitivity than previous space - based missions .
here we focus on studying the 10@xmath0 silicate features of ttss with the infrared spectrograph ( irs ; * ? ? ?
* ) on board spitzer .
it is generally believed that the disks and planetary systems of young stellar objects ( ysos ) form from material from the ism . _
spitzer _ irs spectra of objects with class i spectral energy distributions ( seds ) @xcite , objects believed to be young protostars still surrounded by collapsing envelope material from their parent cloud of gas and dust , show smooth , featureless 10@xmath0 absorption profiles , indicating amorphous silicates .
@xcite presented the spectrum of coku tau/4 , a t tauri star with a 530@xmath5 m spectrum well modeled by @xcite by an accretion disk nearly devoid of small dust grains within @xmath610au . unlike the complex 10@xmath0 emission features of many
herbig ae / be stars indicative of thermally processed silicates @xcite , the 10@xmath0 emission feature of coku tau/4 is smooth , relatively narrow , and featureless , as are the silicate absorption profiles of the ism ( e.g. , * ? ? ?
* ) and class i ysos @xcite .
smooth , narrow , and featureless profiles indicate amorphous silicates .
other objects , such as fn tau @xcite , have significant crystallinity of silicate dust in their disks , evident in the structure in their 10@xmath0 emission features , and others still , such as gg tau a , have larger grains as shown by the greater width of the 10@xmath0 emission feature . in the following , we use optical constants and opacities of various materials to model the 10@xmath0 features of our objects .
we model the dust emission of the six objects whose spectra are presented by @xcite , tw hya and hen 3 - 600 a by @xcite , v410 anon 13 by @xcite , and gm aur and dm tau by @xcite ; we also present and model the 514@xmath0 spectrum of gg tau b. stellar properties for our tts sample are given in table 1 . in 2 , we describe our data reduction techniques . in 3 ,
we detail how we derive and fit an emissivity for each object , and in 4 we describe the fit to the derived emissivity of each object .
we discuss our fits in 5 and summarize our findings in 6 .
the present 12 tts were observed with the irs on board _
spitzer _ over three observing campaigns from 2004 january 4 to 2004 march 5 .
all objects were observed with both orders of the short - low ( sl ) module ( r @xmath660120 ; second order [ sl2 ] , @xmath7@xmath3=0.06@xmath0 [ 5.27.5@xmath0 ] ; first order [ sl1 ] , @xmath7@xmath3=0.12@xmath0 [ 7.514@xmath0 ] ) .
fainter objects were observed with the long - low ( ll ) module ( r@xmath660120 ; second order [ ll2 ] , @xmath7@xmath3=0.17@xmath0 [ 1421.3@xmath0 ] ; first order [ ll1 ] , @xmath7@xmath3=0.34@xmath0 [ 19.538@xmath0 ] ) , while brighter objects were observed with the short - high ( sh ; r@xmath6600 , 9.919.6@xmath0 ) and long - high ( lh ; r@xmath6600 , 18.737.2@xmath0 ) modules ( the lh spectra are not used here ) .
the brightest objects were observed in mapping mode , in which for one module one data collection event ( dce ; sampling of spectral signal from target ) was executed for each position of a 2@xmath93 ( spatial direction @xmath9 spectral direction ) raster centered on the coordinates of the target . for details on how the 2@xmath93 maps were obtained ,
see the description of spectral mapping mode by @xcite . from mapping - mode observations ,
we derive our spectra from the two positions in the 2@xmath93 map for which the flux levels in the raw extracted spectra are highest .
all other objects were observed in staring mode , which always immediately followed single high - precision pointing calibration and reference sensor ( pcrs ) peak - up observations . for details on irs
staring mode operation and pcrs observations , see @xcite . for staring mode ,
the expected flux density of the target determined the number of dces executed at one pointing of the telescope ; for faint objects , multiple dces were obtained at one pointing of the telescope and averaged together . from the position of each target s point - spread function ( psf ) in the cross - dispersion direction in the two - dimensional data ,
we conclude that mispointing in the cross - dispersion direction in sl in mapping mode and in staring mode is usually less than 0@xmath109 ( half a pixel ) . by comparing the absolute flux levels of the spectra obtained from each of the three positions in a 1@xmath93 subsection ( the three positions are colinear and offset from each other in the dispersion direction of the slit ) of the 2@xmath93 raster
, the pointing of the telescope in the dispersion direction in mapping mode could be determined .
the mapping mode dispersion direction mispointing is usually less than half a pixel in sl ( 0@xmath109 ) . pointing in the dispersion direction
can not be quantified very easily for staring mode observations because the telescope is not moved in the dispersion direction in this mode . to account for mispointing in the cross - dispersion direction ,
we use data that has been divided by the flatfield from the s11 pipeline . mispointing in the dispersion direction primarily causes small photometric error ; this will not affect the emissivities derived here .
the spectra were reduced using the spectral modeling , analysis , and reduction tool ( smart ; * ? ? ?
* ) . from basic calibrated data ( bcd ;
flat - fielded , stray - light - corrected , dark - current - subtracted ) s11.0.2 products from the _ spitzer _ science center irs data calibration pipeline , permanently bad ( nan ) pixels were fixed in our two - dimensional spectral data .
the corrected pixel value was a linear interpolation of the nonbad pixels of the set of four nearest neighboring pixels ( up , down , left , and right of the pixel in question ) .
unresolved lines of [ ne ii ] and molecular hydrogen , seen in the spectra of other objects , provided the wavelength calibration ; all spectra presented in this paper are wavelength - corrected , and these wavelengths are estimated to be accurate to @xmath110.02@xmath0 .
all dces taken at the same pointing of the telescope ( same module / order / nod position ) were averaged together . because one order records the spectrum of sky @xmath613away from the target
whose spectrum is being recorded in the module s other order , sky subtraction in low resolution spectra obtained in staring mode is accomplished by subtracting the average spectrum from one spectral order of a given module from that in which the spectrum is located in the same nod position of the other order of the same module . for fm tau ,
the ssc pipeline introduced artifacts in the off - order images in sl2 . in this case
, we used the same - order , different - nod dce to subtract the sky .
the low - resolution sky - subtracted spectra were then extracted using variable - width column extraction in order to account for the linear increase of size of the object s psf with wavelength . from shortest to longest wavelengths of each module , respectively , extraction region width varied from 3.2 to 4.9 pixels in sl2 , 3.6 to 4.4 pixels in sl bonus order ( a short fragment of first - order light from 7.58.4@xmath0 recorded simultaneously with sl2 ) , 2.7 - 5.4 pixels in sl1 , 3.2 - 4.9 pixels in ll2 , 3.8 - 4.3 pixels in ll bonus order ( a short fragment of first - order light from 19.4 to 20.9@xmath0 recorded simultaneously with ll2 ) , and 2.5 to 5.5 pixels in ll1 . for sh the sky is not subtracted , as the sh slit is only 5 pixels long , and no separate sky observations were acquired .
the extraction region at each wavelength for sh was the entire 5 pixel long slit .
since the roughly square sh pixels are @xmath62@xmath102 wide and sl pixels are @xmath61@xmath108 wide , the sh extraction region covered more solid angle at every wavelength than sl , and as spectra from sh typically gave lower flux than sl over the wavelength range of overlap of sh with sl ( @xmath61014@xmath0 ) , the sky levels for the sh observations are estimated to be much lower than the flux density of the point sources .
the spectra are calibrated using a relative spectral response function ( rsrf ) , which gives flux density , f@xmath12 , at each wavelength based on the signal detected at that wavelength .
the rsrfs were derived by dividing the template spectrum of a calibrator star by the result of extraction of the calibrator s spectrum in smart for each nod of each order of each module . for both orders of sl and for both orders of ll , a spectral template of @xmath13 lacertae ( a1v ; m. cohen 2004 , private communication ) of higher spectral resolution than the templates described by @xcite
was used , and a spectral template for @xmath14 dra ( k2iii ) was used for sh @xcite . as with the science targets , sh observations of the calibrator source @xmath14 dra
were not sky - subtracted .
the science target raw extractions were then multiplied by the rsrfs corresponding to the same nod , order , and module . typically , good flux agreement at wavelength regions of order overlap within the same module was found . as the spectra obtained for a given source at the telescope s two nod positions are independent measurements of the object s spectrum
, close agreement between the two nod positions was expected ; this was the case for all sources except for fn tau .
fn tau was observed in mapping mode , and from the extracted flux levels of all observations in the 2@xmath93 mapping raster , we determined that the central 1@xmath92 pair both suffer mispointing of differing amounts .
the more mispointed dce was mispointed in the dispersion direction by 0@xmath107 , while the less mispointed dce was mispointed in this direction by 0@xmath103 ; both dces were mispointed in the cross - dispersion direction by between 0@xmath105 and 0@xmath107 .
the effect of this differential mispointing shows up most prominently in the derived spectrum of the more mispointed observation of first order of sl ; there is a mismatch of flux level of about 10@xmath15 over the entire order compared to the flux level of the spectrum obtained from the less mispointed dce . to correct for this ,
the first order of the spectrum from the more mispointed mapping position was multiplied by 1.1 to match the less mispointed position . except for the first order of sl for fn tau , the derived spectra are the mean at each order of each module of the spectra from the two independent nod or map positions .
for the first order of sl for fn tau , the reported spectrum is that from only the less mispointed map position .
error bars are derived for each of the spectra , and the error bar at a given wavelength is equal to half the difference between the flux ( at that wavelength ) from the two nod ( or map ) positions used to derive the mean spectrum . for sl first order of fn tau
, the spectrum from the less mispointed map position and the corrected spectrum ( previously described ) from the more mispointed map position were used to derive its error bars . any error bar with relative uncertainty
@xmath11% is attributed to the low number ( 2 ) of measurements at that wavelength , and that errorbar is set to 1% of the flux .
there are some mismatches in flux between sl and sh , and between sl and ll in the spectra . comparing the sl spectra of nonvariable sources to available photometry
, absolute spectrophotometric accuracy is estimated to be better than 10% in sl .
therefore , small mismatches in flux levels between sl , sh , and ll are corrected by scaling the entire longer wavelength module to match the flux in sl , as we trust the photometric levels of sl .
sh was multiplied by factors between 1.04 and 1.11 to match sl ; ll for cy tau was multiplied by 0.95 to match its sl spectrum . in order to account for off - order leaks in the filters which define the orders of each of the modules ,
the ends of each order of every module are truncated to guarantee the spectral purity of our spectra .
the spectra of all objects in the sample excluding gg tau b have been previously published : coku tau/4 , fm tau , ip tau , gg tau a , fn tau , and cy tau in @xcite ; tw hya and hen 3 - 600 a in @xcite ; v410 anon 13 in @xcite ; and gm aur and dm tau in @xcite .
for all previously published spectra except the two by @xcite , wavelengths from 8 to 14@xmath0 were too long by 0.05@xmath0 ; as described previously , this wavelength problem was corrected before further analysis .
the correction has moved the 9.4@xmath0 feature in fn tau noted by @xcite closer to 9.3@xmath0 . in figure 1 ,
the spectrum obtained of the gg tau b binary system is shown ; in 4 we discuss the origin of the ir excess for this pair .
for all of the objects in the sample except v410 anon 13 , no correction for extinction is applied , in order not to introduce artifacts of overcorrection for extinction .
see table 1 for the assumed visual extinction @xmath16 for each of the objects in our sample .
no extinction correction is applied for any object having @xmath16 less than 1.4 ; this includes all objects in the sample except coku tau/4 , gg tau a , and v410 anon 13 . as described by @xcite , optical spectra of coku tau/4
indicate time - dependent reddening to the star .
this suggests that the source of the extinction to coku tau/4 may be local to the object .
however , the precise time and space dependence of extinction to coku tau/4 is unknown .
for this reason , no extinction correction is applied for coku tau/4 .
@xcite estimate @xmath16 of @xmath4 3.2 toward gg tau ab ( the gg tau system is a hierarchical quadruple with the northern pair , gg tau a , being binary and separated by 0@xmath1025 this is described in greater detail in 4 ) , while @[email protected] toward gg tau aa .
consequently , no extinction correction is applied for gg tau a , as it is believed the less extinguished gg tau aa component dominates the irs spectrum . for v410 anon 13 ,
an extinction correction is applied , assuming @[email protected] along with @xcite .
furlan present a disk model for this object to fit its irs spectrum assuming an inclination @xmath18 = 70 . because of the large implied disk inclination , at least part of the extinction to v410 anon 13 could be due to dust in the flared disk atmosphere at large disk radii lying in the sightline from the star and inner disk regions to observer .
@xcite also found that when @xmath18 in the model is changed from 70to 60 , the peak of the flux in the 10@xmath0 feature increases by @xmath620@xmath15 . because the emergent disk spectrum in the model greatly depends on its inclination @xmath18 , the effect of extinction correction for this object
is discussed in 4.12 when describing its dust model fit . because dust in the outermost reaches of yso disks
is expected to be little altered from its origin in the ism ( see the discussion in 5 below ) , the composition of any dust providing local extinction is assumed to be approximately the same as that in the ism between v410 anon 13 and earth . in the ism ,
the ratio of the visual extinction to the optical depth at the 9.7@xmath0 peak of the silicate absorption feature ( @xmath16/@xmath7@xmath19 ) varies by as much as a factor of 2 to 3 ( see draine 2003 ) . to convert from @xmath16 to the 9.7@xmath0 extinction ,
we take @xmath16/@xmath7@xmath19@xmath1718 , typical for the local diffuse ism ( see draine 2003 ) . for simplicity , we assume that the composition of the material responsible for the extinction does not change over the sightline from the target to earth . the spectral excess for each of the objects in the sample is interpreted as arising from a disk surrounding one or more central star(s ) beginning in most cases at a few stellar radii away from the central star(s ) and extending as far away as a few hundred au .
we call a disk a `` transitional disk '' if it is optically thick to mid - ir wavelengths over some range of radii and optically thin elsewhere .
coku tau/4 @xcite , dm tau @xcite , gm aur @xcite , tw hya @xcite , and hen 3 - 600 a @xcite have been shown to be transitional disks through spectral modeling .
the spectrum of coku tau/4 is photospheric at wavelengths shortward of 8@xmath0 but has a large ir excess seen in its irs spectrum longward of that wavelength ; correspondingly , it has an optically thick disk at radii greater than 10au with less than @xmath60.0007 lunar masses of small silicate dust grains inside that radius @xcite .
dm tau has an irs spectrum similar to that of coku tau/4 and is modeled similarly by @xcite , but with the radius of transition between the optically thick disk and the ( very ) optically thin inner regions at 3au .
gm aur , tw hya , and hen 3 - 600 a , also with large excess above photosphere longward of 8@xmath0 , are not photospheric shortward of 8@xmath0 .
this excess indicates an optically thin inner disk region .
the ir disk emission for each of the transitional disks is isolated by subtracting an appropriate stellar photosphere represented by a blackbody . for coku tau/4 and dm tau ,
the blackbody is fit to the 58@xmath0 irs spectral data , while for gm aur , tw hya , and hen 3 - 600 a the rayleigh - jeans tail of the blackbody is fit to that of the stellar photosphere model by @xcite and @xcite , respectively .
this isolated disk emission for each object should , shortward of @xmath620@xmath0 , be due in large part to emission from the optically thin regions of the disk .
for coku tau/4 and dm tau , this emission is mostly due to the optically thin regions of each object s wall . for gm aur , tw hya , and hen 3 - 600 a , this emission is mostly due to the optically thin inner disk regions @xcite .
we refer to a `` full disk '' if the disk is optically thick to mid - ir wavelengths throughout and extends from the dust - sublimation radius from the central star . following the reasoning of @xcite , fm tau , ip tau , gg tau a , gg tau b , fn tau
, v410 anon 13 , and cy tau are identified as having full disks based on their 58@xmath0 spectra .
each has a continuum from 5 to 8@xmath0 characterized by a spectral slope shallower than the rayleigh - jeans tail from a naked stellar photosphere .
in addition , the 5 to 8@xmath0 flux exceeds that from stellar photosphere alone ( modeled by fitting a stellar blackbody to the near - ir photometry ) by factors @xmath202 .
following the discussion by @xcite , most of the 58@xmath0 emission from full disks originates from optically thick inner disk regions , while most of the emission in the dust features above the continuum longward of 8@xmath0 is due to emission from dust suspended in the optically thin disk atmosphere .
therefore , a power law continuum is fit to the @xmath18@xmath0 region of each `` full - disk '' spectrum and subtracted from the spectrum to isolate the optically thin disk atmospheric emission .
dust grains suspended in the optically thin atmosphere of a flared disk are directly exposed to stellar radiation , which heats the grains above the temperature of the disk s photosphere .
the grains then reemit the absorbed energy according to their temperature ; this emission gives rise to the distinctive dust features seen in the spectra beyond 8@xmath0 .
the emission features are much narrower than a planck function , which indicates structure in the dust emissivity .
as explained in @xcite , a radiatively heated disk with a modest accretion rate has a thermal inversion .
the upper layers of the disk , the optically thin disk atmosphere , are hotter than the lower layers , which are optically thick .
this gives rise to spectral emission features characteristic of the dust in the disk atmosphere . in modeling the seds of classical t tauri stars ( cttss ) , both @xcite and @xcite compute the temperature of the atmosphere of each annulus of disk material as a function of vertical optical depth . in such models , it is assumed that all dust grains at a given height in an annulus are at the same temperature , independent of grain composition and grain size .
it is similarly assumed here that all grains in any sufficiently small volume in a disk are at the same temperature , independent of grain composition and grain size .
we aim for a simple model in order to determine the composition of the part of the disk giving rise to the optically thin dust emission . @xcite and @xcite
also model dust emission by assuming a single temperature for all dust components .
optically thin emission from dust over the range of radii ( and therefore temperatures ) that contributes most to the 820@xmath0 range is represented by emission from optically thin dust at a single , `` average '' temperature , @xmath21 .
it is assumed that the monochromatic flux of this optically thin emission over the short range from 8 to 14@xmath0 is given by @xmath22 where @xmath23 is either the photosphere - subtracted residuals ( for transitional disks ) or power - law - continuum - subtracted residuals ( for full disks ) ; @xmath24 is referred to as the emissivity ; @xmath25 is the solid angle of the region of optically thin emission ; and @xmath26 is the frequency - dependent optical depth of dust . for all objects except gg tau b and gm aur , @xmath21
is found by assuming a long- to short - wavelength emissivity ratio , @xmath27/@xmath28 , with `` @xmath29 '' meaning long wavelengths ( @xmath620@xmath0 ) and `` @xmath30 '' meaning short wavelengths ( @xmath610@xmath0 ) , for dust and solving for @xmath21 in the equation @xmath31 for gg tau b , where no long - wavelength data exist , a dust temperature of 252k is assumed , the same temperature as for gg tau a. for reasons discussed in 4.4 , the dust temperature is set to @xmath21=310k for gm aur . for all other objects observed with sl and ll excluding dm tau , we take the 20@xmath0 - 10@xmath0 flux ratios ; for the objects observed in sl , sh , and lh , we decrease the wavelengths in the ratio to 19.3 and 9.65@xmath0 , as sh does not extend to 20.0@xmath0 . in a single - temperature dust model , the same temperature will be computed regardless of the wavelengths used to determine the flux ratio due to the properties of the planck function .
the wavelengths were set to 9.5 and 19.0@xmath0 for dm tau as a test to determine if the derived temperature depended much on the exact choice of wavelengths used to the determine flux ratio . when the wavelengths were changed to 10 and 20@xmath0 for dm tau , the computed temperature changed from 160 to 158k ; however , this did not require any change to the dm tau dust model .
a similar test was performed on fn tau by changing the long and short wavelength fluxes used for its dust temperature determination from 10 and 20@xmath0 to 9.5 and 19@xmath0 .
this increased dust temperature from 208 to 209k ; as with dm tau , no change to the fn tau dust model was required .
using this temperature , @xmath21 , the photosphere- or continuum - subtracted residuals were divided by @xmath32 to give the emissivity , which is proportional to the mass - weighted sum of opacities as follows : @xmath33 where @xmath34 is the mass fraction of dust component @xmath35 , @xmath36(@xmath3 ) is the wavelength - dependent opacity ( @xmath37g@xmath38 ) of dust component @xmath35 , and @xmath39(@xmath3 ) is the wavelength - dependent cross - section of the dust mixture model . both @xmath40(@xmath3 ) and @xmath39(@xmath3 )
are normalized to unity at their peak in the 814@xmath0 range . to determine the uncertainties in the emissivities , the corresponding spectral error bars , obtained as described in 2.3 from spectra obtained at two nod positions , are divided by @xmath32 , the result of which
is then divided by the same normalization constant used to derive the corresponding emissivity .
the derived emissivities are believed to be valid immediately longward of 8@xmath0 , where the 10@xmath0 feature rises above the extrapolation of the @xmath18@xmath0 continuum , as no drastic change of the slope of the continuum from the optically thick components of the disks at wavelengths between 8 and 14@xmath0 is expected . however , assuming one dust temperature for a wide range of wavelengths in a spectrum of a circumstellar disk is unrealistic .
in addition , for wavelengths longward of @xmath614@xmath0 , the slope of the continuum from optically thick emission is not well determined by extrapolation from the @xmath18@xmath0 continuum .
the power - law - continuum- or photosphere - subtracted residual flux at @xmath620@xmath0 , attributed as described previously to optically thin emission , is therefore uncertain , leading to uncertainty in the derived dust temperature .
for this reason , we do not attempt to fit the 18@xmath0 and longer wavelength features here .
the emissivity is fit by finding the optimal set of mass fractions , @xmath34 , such that the normalized model dust cross - section fits the normalized emissivity as well as possible .
the fitting method is iterated until the assumed @xmath27/@xmath28 , used to compute grain temperature and therefore derive emissivity , equals the @xmath27/@xmath28 derived from the fit emissivities ; @xmath27/@xmath28 and other details of derivation of emissivities are reported for each of the opacity models in table 2 .
also listed in table 2 is @xmath41 , the ratio of the continuum - subtracted residual flux at 9.9@xmath0 to the 9.9@xmath0 continuum of the full disks , which gives the contrast of the silicate emission feature to the optically thick continuum in the original spectrum . as a test of these simple dust models , a disk model following the methods of @xcite
was computed , using stellar parameters from table 1 and the mass accretion rate from @xcite .
first , opacities similar to those generated from the fit to the emissivity of ip tau ( see 4.7 ) were used to determine the temperature structure of the disk , but it was found that the disk would not heat up sufficiently the spectrum generated from this model had lower flux than the ip tau spectrum at all wavelengths . a source of dust opacity with high absorption at visible and near - ir wavelengths was needed to absorb the stellar radiation and heat up the disk to give more flux at all irs wavelengths ( 536@xmath0 ) .
facing the same issue when attempting to model the interstellar extinction , @xcite added in an artificial source of absorption for @xmath3@xmath18@xmath0 in their `` astronomical silicates '' for similar reasons ( also see jones & merrill 1976 ; rogers 1983 ) .
therefore , the optical properties of graphite and `` astronomical silicates '' from @xcite were used to determine the radial and vertical temperature structure of the disk .
then , opacities obtained from the fit to the ip tau 10@xmath0 feature were used to determine the emergent spectrum .
this disk model uses the same dust opacities at all radii and at all heights in the disk . as can be seen in figure 2 ,
the 512@xmath0 model spectrum is in quite good agreement with the observed spectrum of ip tau .
it is concluded that the emissivities derived here are basically correct . at wavelengths
@xmath2012 @xmath5 m , the model gives more flux than is observed .
there is too much emission from cool grains with respect to that from warm grains .
this may be an indication that dust in the ip tau disk is more settled ( settling meaning that the dust - to - gas ratio in the disk atmosphere is lower than that at the disk midplane ) than assumed in the model . according to @xcite and @xcite , more
settling of dust reduces disk emission at wavelengths @xmath2020@xmath0 more than shortward of 20@xmath0 .
the material in yso circumstellar disks originates from the ism .
@xcite fit the interstellar extinction with silicates and graphite ; here , the spectra are also fit with opacities derived from silicates and carbonaceous material .
silicates , which contain sio@xmath42 groups , give rise to 10 and 18@xmath0 features from the bonds between si and o. the 10@xmath0 feature arises from si o stretching modes , while the 18@xmath0 feature is due to o si o bending modes .
crystalline silicate grains have many narrow , strong resonances , while amorphous silicate grains , lacking ordered structure , give rise to two broad , weaker resonances ; one near 10@xmath0 and one near 18@xmath0 . from @xcite ,
the optical constants for amorphous pyroxene of composition mg@xmath43fe@xmath44sio@xmath45 and amorphous olivine of composition mgfesio@xmath46 are used . in this paper , these dust components are referred to as `` amorphous pyroxene '' and `` amorphous olivine , '' respectively ; collectively , they are referred to as `` amorphous silicates . '' for optical constants of the three crystalline axes of forsterite of composition mg@xmath47fe@xmath48sio@xmath46 , data from @xcite is used .
the same bulk density , 3.3g@xmath49 , is assumed for forsterite and all small amorphous olivine and pyroxene grains .
this value is obtained from @xcite for the high end of the range of densities for `` high - t silicate '' . for large amorphous olivine and pyroxene grains ,
this value is the bulk density of the matrix for large porous amorphous olivine and pyroxene grains ( see 3.5 for discussion on how porosity modifies the density of a grain ) . for crystalline pyroxene , mass absorption coefficients for mg@xmath50fe@xmath48sio@xmath45 ( en90 , having mg/(mg+fe ) @xmath4 0.9 ) by @xcite are used .
chihara report this crystalline pyroxene to have monoclinic structure . to represent the opacity profile of silica ( sio@xmath51 ) ,
the dispersion parameters for the ordinary ray and extraordinary ray of @xmath13 quartz by @xcite are used .
the density of @xmath13 quartz assumed is 2.21g@xmath49 , which is the value for amorphous silica stated by @xcite .
this is also close to the values of 2.27 and 2.32g@xmath49 for tridymite and cristobalite ( polymorphs of silica ) , respectively @xcite .
such a value for the density of quartz is chosen because it is deemed likely ( see discussion in 4 ) that forms of silica other than @xmath13 quartz are present in the disks studied here , possibly in greater mass fractions than @xmath13 quartz .
the nominal value for @xmath13 quartz is 2.65g@xmath49 , as is given in @xcite . because opacity is inversely proportional to density , and
because the density of quartz is potentially underestimated here , the silica profile may need to be multiplied by a constant slightly less than 1 ; therefore , our mass fractions for quartz could be underestimated by as much as a factor of ( 2.21/2.65)@xmath60.83 . for the t tauri sample in this paper
, we find that adequate fits to the dust emissivity do not require large crystalline silicate grains .
as articulated by @xcite , glassy amorphous silicates produced and analyzed in laboratories on earth are too `` clean '' , meaning that they do not give enough opacity in the 1316@xmath0 trough between the 10 and 18@xmath0 silicate features to account for the dust emissivities of mira variable stars .
clean silicates also have too little absorption shortward of 8@xmath0 ; jones & merrill ( 1976 ) and rogers ( 1983 ) argued that circumstellar silicates must be `` dirty '' to account sufficiently for the absorption of stellar radiation by dust in shells around evolved stars .
the derived disk emissivities in our study of t tauri stars also require a relatively featureless source of continuum opacity longward of 13@xmath0 .
the source of this continuum opacity is not well constrained because it has no distinctive spectral features .
we use amorphous carbon to supply the continuum opacity , as carbonaceous material is believed to be a major component of the ism ( e.g. , * ? ? ?
* ; * ? ? ?
@xcite argue in favor of amorphous carbon as a major component of interstellar and circumstellar dust , and @xcite asserts that a large fraction of the carbon found in chondritic interplanetary dust particles ( idps ) is amorphous . for amorphous carbon ,
the optical constants for the `` acar '' mixture @xcite are used , and a bulk density for amorphous carbon grains of 2.5g@xmath49 @xcite is assumed .
in addition to grain composition , grain shape and grain size affect the computed dust opacities .
the presence of interstellar polarization indicates that interstellar grains must be both elongated and aligned ( draine 2003 ) .
@xcite observed in the becklin - neugebauer ( bn ) source in orion a correlation of the linear polarization with the optical depth of the 10@xmath0 silicate band , leading them to propose the existence of nonspherical aligned silicate grains in front of bn . for an approximation of actual grain shapes , we assume here a cde2 shape distribution @xcite for small amorphous silicates and amorphous carbon dust in our models . in cde2 ,
all ellipsoidal shapes are included , but the distribution is peaked toward near - spherical shapes , and extreme shapes such as needles or flat sheets are given zero weight .
cde2 requires all grains to be in the rayleigh limit ( 2@xmath52a/@xmath3 @xmath53 1 , where a is the radius or characteristic size of grain ) .
this condition is generally met by submicron - sized grains for the irs wavelengths ( @xmath3@xmath205@xmath0 ) . for the shape distributions of quartz and forsterite , the continuous distribution of ellipsoids ( cdes ;
* ) is used , in which all shapes are equally weighted , because this shape distribution fits the observed dust features in the emissivities better than the cde2 shape distribution .
we note in the appendix that 0.1@xmath0 sized porous grains of @xmath13 quartz and forsterite give rise to opacity profiles nearly identical to those of solid grains of @xmath13 quartz and forsterite with the cde distribution , respectively .
we do not suggest that the small crystalline grains are porous , but we note this identity . for forsterite , it is assumed that the three crystalline axes are randomly oriented with respect to the three ellipsoidal axes .
as with cde2 , cde also assumes particles much smaller than the wavelengths of interest .
because opacities for crystalline pyroxene grains ground in an agate mortar @xcite are used , no definitive statements about the grain shape distribution for this species of dust can be made .
as described by @xcite , forsterite was ground into powder ; when observed with a scanning electron microscope , the forsterite particles constituting the powder are observed to be nonspherical .
the only statement that can be made about these particles sizes is that they are submicron @xcite .
the same is assumed to be true for the crystalline pyroxene grains described by @xcite .
grain growth is accounted for by including opacities derived from grains of a single size to represent large grains , following @xcite and @xcite ; see those references for descriptions of the sizes and shapes of the grains they used .
mie theory @xcite can be used to obtain exact results for the opacity of solid spherical grains of arbitrary size ; however , large grains grown by coagulation in protoplanetary disks are expected to have an irregular and porous structure ( see li & lunine 2003 and references therein ) . to account for this heterogeneity of large grains , the bruggeman effective medium theory ( emt ; bohren & huffman 1983 ) is used , following li & lunine ( 2003 ) , in deriving the effective dielectric functions of large fluffy grains .
@xcite found that using these effective dielectric functions in mie theory could approximate the absorption efficiency of grains with fractal dimension @xmath54 , a measure of the porosity ( volume fraction of empty space ) in a given grain , between 2.5 and 3 ; @xmath54=3 for solid grains , while @xmath54@xmath13 represents porous grains .
the effective dielectric functions computed from the bruggeman emt are used in mie theory to derive the absorption efficiencies of spherical porous grains . @xcite
use the emt with 2.5@xmath1@xmath54@xmath13 to model the fluffy grains in comet hale - bopp ( also see li & greenberg 1998 ;
lisse et al .
the optical depth through the center of the grains is preserved , in order to match the opacity profile of the solid 2@xmath0 amorphous silicate grains used by @xcite . optical depth ( @xmath55 ) and bulk silicate opacity ( @xmath56 )
are kept the same for both solid and porous grains , assuming the silicate matrix of a porous grain has the same opacity as that of an equivalent - mass solid grain , and the diameter for the porous grain is obtained by @xmath55=@xmath57 , where @xmath58 is the mass density of the grain [ equal to the bulk silicate density ( that of a solid silicate grain ) times @xmath59 , where @xmath60 is the volume fraction of vacuum ] . for @xmath60=0.6 , a porous grain of radius 5@xmath0
has the same optical depth as that through a solid 2@xmath0 radius sphere .
the fractal dimension was chosen to be 2.766 , a number which is within the acceptable range for @xmath54 .
the 814@xmath0 opacity profile of these porous grains with radii @xmath61=5@xmath0 , @xmath54=2.766 and @xmath60=0.6 ( i.e. 60@xmath15 of their volume is vacuum ) is very similar to that of solid spherical ( @xmath60=0 , @xmath54=3 ) amorphous silicate grains of radii @xmath172@xmath0 . in this paper , by `` large grains '' we mean porous amorphous olivine or pyroxene grains of radii @xmath61=5@xmath0 , volume fraction of vacuum @xmath60=0.6 , and dimension @xmath54=2.766 .
our fits to the disk emissivities are not necessarily unique .
the degeneracy in the mass fraction of amorphous silicate components is larger than for crystalline silicate components .
for instance , much of the broad , smooth opacity profile of either of the large silicate grain components overlaps with both that of the other large silicate grain component and both of the small amorphous silicate opacity profiles .
there is less uncertainty associated with the two small amorphous silicate profiles , as they are narrower and stronger than the profiles of the large amorphous silicates .
there is even less uncertainty regarding the crystalline silicate components , as their profiles are even narrower and often stronger than any of the amorphous silicate profiles .
some crystalline silicate identifications may be confused by overlapping bands : the opacity profile of crystalline pyroxene has a peak at @xmath69.3@xmath0 that is close in wavelength to the 9.2@xmath0 feature of quartz ; the prominent @xmath611.3@xmath0 peak of forsterite is very close to the weak 11.2@xmath0 feature of crystalline pyroxene . in addition , there are polymorphs and forms of silica other than @xmath13 quartz ( see 3.4 ) .
@xcite gives 7.513.5@xmath0 extinction profiles for silica ; they all peak at around 9.19.2@xmath0 .
this implies a very large degeneracy between the different polymorphs and forms of silica .
in addition , there is degeneracy between different forms and polymorphs of crystalline pyroxene .
the extinction efficiency of clinopyroxene from @xcite shows a 9.3@xmath0 peak higher than the other two major peaks in the 10@xmath0 region .
if one takes the @xcite profile of crystalline pyroxene ( en90 , which is monoclinic ) and adds silica , boosting the 9.3@xmath0 feature , and also forsterite , boosting the crystalline pyroxene features between 11 and 12@xmath0 , one obtains a profile with feature strengths similar to that reported for clinopyroxene in @xcite .
therefore , there is degeneracy between clinopyroxene and a mixture of silica , forsterite , and the crystalline pyroxene used here ( en90 ) .
the featureless opacity continuum from amorphous carbon may be reproduced by silicate grains larger than @xmath620@xmath0 in radius , which give essentially featureless continuum for 820@xmath0 wavelengths .
the 10@xmath0 emissivity profiles of the ism , @xmath5 cep , and three of our transitional disks are compared in figure 3 , as described in 4.1 and table 3 . in figures 4 , 5 , and 6
the derived emissivities and model fits are shown for the 12 t tauri stars .
the parameters of the model fits are detailed in table 4 .
figure 3 compares the smoothest , simplest 10@xmath0 silicate emission features in the tts sample to the derived optical depth profile of the interstellar absorption towards gcs3 in the galactic center from ( * ? ? ?
3b ) and the 10@xmath0 feature of @xmath5 cep @xcite , a mass - losing m2 ia supergiant @xcite . according to @xcite , the visual extinction @xmath16 for @xmath5 cep is @xmath62 , so no correction for extinction is applied . for @xmath5 cep ,
the emissivity was derived by first subtracting a model photosphere of @xmath62=3500k , log(g)=0 , and solar metallicity by @xcite , binned to the _ iso _ short wavelength spectrometer ( sws ) spectral resolution . following the procedure outlined in 3.2 , a silicate temperature of 423k
was derived , based on the residuals at 10 and 18.2@xmath0 . the gcs3 profile from @xcite
is displayed unaltered . in order to compare the tts and @xmath5 cep profiles with the gcs 3 profile , a first - order baseline was fit to each of the emissivities at @xmath68 and @xmath613@xmath0 and subtracted from the corresponding emissivity ; each residual
is scaled to the peak of the gcs 3 profile .
the parameters used for this process are listed in table3 .
all five silicate profiles in figure3 are generally smooth .
the profiles of tts peak at wavelengths closer to the peak of the ism profile than to that of @xmath5 cep , which peaks at a somewhat longer wavelength .
the close similarity between the gcs3 ism profile and the tts profiles strongly supports the assertion that the dust in yso disks comes from the ism .
the smoothest , least complex t tauri 10@xmath0 feature is that of coku tau/4 .
@xcite model this object as a standard flared disk of modest inclination with nearly all small grains in the inner 10au cleared : literally , a disk in transition . this modeling indicates less than 0.0007 lunar masses of grains of ism grain size ( @xmath10.25@xmath0 ) remain in the inner disk .
this configuration is remarkable considering the estimated age of the coku tau/4 system , which is only 13myr .
the 10@xmath0 silicate feature in this model is dominated by emission from the cylindrical disk `` wall '' at the 10au truncation radius .
the outer layers of this wall are roughly isothermal , as each point on the wall is @xmath610 au from the star ; as such , the emissivity for coku tau/4 , derived by assuming all grains are at a single temperature , is a reasonable first approximation .
the derived dust temperature for coku tau/4 is @xmath6121k , which is well within the range of temperatures in figure 4 of @xcite computed using amorphous pyroxene and amorphous olivine grains in the optically thin region of the coku tau/4 wall .
this temperature alone indicates the bulk of the material is located @xmath610au from the star as noted by @xcite .
the 10@xmath0 silicate profile of coku tau/4 is quite similar to that of the ism ( see figure 3 ) . a satisfactory fit ( figure 4 )
is achieved with nonspherical small amorphous pyroxene and olivine grains with a mass ratio of 3.73:1 .
@xcite fit the 10@xmath0 interstellar silicate absorption profile toward the galactic center source gcs3 with small spherical amorphous pyroxene and olivine grains with a mass ratio of 0.18:1 .
most of this difference in mass ratio comes from the differing adopted shape assumptions .
the cde2 grain shape distribution assumed here shifts the 9.8@xmath0 peak from spherical amorphous olivine grains to 9.95@xmath0 and the 9.3@xmath0 peak from spherical amorphous pyroxene grains to 9.4@xmath0 . in addition , the coku tau/4 profile peaks at a slightly shorter wavelength ( 9.55@xmath0 ) than the gcs3 ism profile ( 9.6@xmath0 ) .
this could indicate a slight compositional difference , although note the large error bars for @[email protected]@xmath0 in coku tau/4 .
the lack of narrow features on top of the amorphous feature indicates negligible amounts of crystalline grains . at 13 myr ,
the dust from 10au outward in the coku tau/4 disk shows little evidence for processing .
although small nonspherical grains of amorphous olivine and pyroxene account successfully for the peaks of the 10 and 18@xmath0 features , the model with small silicate grains does not account for a very small excess of emissivity on the long - wavelength side of the coku tau/4 10@xmath0 feature .
optically thick emission from the dust would widen the 10@xmath0 feature , but it would do so to both the long- and short - wavelength sides of the feature .
larger amorphous silicate grains , however , can account for this 11.512.5@xmath0 excess .
larger grains ( see 3 ) give greater opacity longward of the 10@xmath0 silicate peak , but not shortward . consequently , porous silicate grains of 5@xmath0 radius are included 4.3@xmath15 by mass in the coku tau/4 dust model .
while models consisting of only silicates can account for the positions and shapes of the 10 and 18@xmath0 features , the 1215@xmath0 opacity continuum from laboratory silicates is lower than that in the derived coku tau/4 emissivity .
amorphous carbon is used to model this excess continuum opacity ( 3 ) .
dm tau is also a disk in transition . at disk radii interior to 3au ,
less than 0.0007 lunar masses of submicron - sized grains remain ( similar to coku tau/4 ) , according to @xcite .
the derived dust temperature for dm tau , 160k , is only slightly higher than that for coku tau/4 ( 121k ) , which reinforces the conclusion that dm tau has no inner dust disk ( see 4.2 ) . as with coku tau/4 , the disk at radii larger than 3au
is optically thick , and the spectrum from 5 to 8@xmath0 is photospheric .
the 10@xmath0 feature is of somewhat low contrast to the underlying photosphere , but our derived emissivity is well fit with small grains of amorphous olivine and pyroxene , large 5@xmath0 porous amorphous olivine grains , and amorphous carbon .
no crystalline grains are indicated by the dm tau profile ( figure 4 ) .
@xcite assert that gm aur is also a transition disk , albeit with a slightly more complex radial distribution of dust than for coku tau/4 or dm tau .
their model includes an optically thin inner disk from the dust sublimation radius out to 5au containing @xmath60.02 lunar masses of small dust grains , negligible small dust grains between 5 and 24au , and a full outer disk beyond 24au .
the usual method of computing dust temperature self - consistently based on the short- to long - wavelength flux ratio derived from the photosphere - subtracted residuals results in a very poor fit to the derived gm aur emissivity .
that method gives @xmath63210k , with 50% by mass small amorphous pyroxene , 33% small amorphous olivine , and 17% amorphous carbon , resulting in a @xmath64 per degree of freedom ( dof ) of @xmath6100 , severely underestimating the emissivity for @xmath3@xmath19@xmath0 and slightly overestimating the emissivity longward of 13@xmath0 .
therefore , a self - consistent dust temperature of @xmath21=386k was determined for gm aur using the fluxes at 9.4 and 18.8@xmath0 of the optically thin inner disk component from the gm aur model of @xcite .
this gave a somewhat better fit to the computed emissivity .
the model required @xmath653% by mass small amorphous olivine and @xmath447 @xmath15 amorphous carbon and gave a @xmath64/dof of 44.4 , with the dust model overestimating the emissivity shortward of the 10@xmath0 feature peak and underestimating the emissivity longward of 12.5@xmath0 .
the low dust temperature of 210k resulted from including emission from the very cold wall at 24au in the model of @xcite .
the second method excludes the wall emission and gave a dust temperature of 386k .
the dust model assuming dust at 210k underestimates the 8@xmath0 emissivity while the model assuming 386k dust overestimates the 8@xmath0 emissivity ( and the opposite effects happen at @xmath614@xmath0 ) .
therefore , the dust temperature is set between 210 and 386k to achieve the optimal fit to the derived emissivity . by setting @xmath21=310k , a @xmath64/dof of @xmath69.8
was achieved using the dust mass fractions listed in table 4 , adequately fitting the computed 814@xmath0 emissivity .
amorphous olivine is included to match the @xmath4 9.8@xmath0 central wavelength of the gm aur 10@xmath0 feature .
amorphous pyroxene and amorphous carbon are also included ; however , no large amorphous silicate grains are indicated , as any amount of such grains would make the model 10@xmath0 feature wider than that observed .
likewise , the spectrum indicates negligible amounts of crystalline silicates and quartz .
tw hya is another suspected transitional disk @xcite . @xcite
propose that it is a disk partially cleared out to @xmath64au ; gas and @xmath60.5 lunar masses of grains of radii @xmath61@xmath0 populate the inner optically thin disk .
@xcite found a transition radius of 3.3au instead of 4au based on the irs spectrum of tw hya .
the outer disk is assumed to be optically thick .
the derived dust temperature for tw hya , 193k , is low ( as with coku tau/4 ) compared to the dust temperatures of other stars in our sample , which further supports tw hya relatively lacking inner disk material .
the 10@xmath0 feature of tw hya is smooth , like those of coku tau/4 and dm tau ; little crystalline material is indicated .
the derived emissivity peaks at 9.55@xmath0 , indicating a high - mass fraction of amorphous pyroxene . a fairly substantial large grain content
is indicated , @xmath624.5% by mass .
approximately 1% by mass of forsterite is indicated by the slight knee at 11.3@xmath0 .
tw hya shows two small peaks at 12.4 and 12.8@xmath0 . according to @xcite , various forms of amorphous and crystalline silica have features peaking between 12 and 13 @xmath0 .
quartz is the only one with two peaks at roughly 12.4 and 12.8@xmath0 .
this indicates @xmath61% quartz by mass . at wavelengths shorter than 8.7@xmath0 , however , the emissivity rises above the model .
the data in @xcite shows that other forms of silica may better fit this short - wavelength `` shoulder . ''
we intend to investigate this in a future study using a larger sample of tts spectra . with an age of @xmath610myr @xcite
, tw hya has very little crystalline material , which is somewhat unexpected .
a number of the disks in taurus from our sample , which like coku tau/4 are believed to be 13myr old , have significantly greater crystalline mass fractions than tw hya .
hen 3 - 600 a , another system in the tw hydrae association , has much greater mass fractions of crystalline grains , making it an interesting counterpoint .
this issue is discussed further in 5 .
fm tau has a slightly more complex 10@xmath0 feature than the taurus transitional disks .
the 58@xmath0 continuum for this object is shallower than , and well in excess of , the stellar photosphere , implying that the optically thick disk around this object extends inward to the dust sublimation radius @xcite .
the 10@xmath0 feature of fm tau peaking around 9.69.7@xmath0 is narrow .
it is fit with negligible amounts of large grains .
forrest et al .
( 2004 ) note that the spectrum of this object has a `` knee '' at 11.3@xmath0 , indicating the presence of forsterite .
the mass fraction of forsterite used in the dust opacity model for fm tau is only @xmath60.3% .
the model spectrum is higher than the emissivity at @xmath18.3@xmath0 .
this is attributed to uncertainty in the subtraction of the power law ( which was fit to the optically thick emission immediately shortward of 8@xmath0 ) , as the derived emissivity immediately longward of 8@xmath0 is more sensitive to the absolute level of the power law than the emissivity at longer wavelengths .
ip tau is similar to fm tau .
@xcite also ascribe a full disk extending to the dust sublimation radius for this object .
the 10@xmath0 feature peaks at @xmath69.69.7@xmath0 and is only slightly wider than that of fm tau , indicating large amorphous olivine grains of @xmath67% by mass .
the 11.3@xmath0 knee of ip tau is more prominent than that of fm tau , indicating a higher crystalline mass fraction of forsterite ( 2% ) than for fm tau .
the opacities used to fit the emissivity of ip tau are very similar to those used in the ip tau disk model described in 3.3 ; there are exceptions . instead of using cde2 as the shape distribution for small amorphous olivine and amorphous pyroxene , the cde shape distribution was used . for the same reasons that the assumed shape distribution affected the ratio of amorphous pyroxene to amorphous olivine for gcs 3 and coku tau/4 ( see discussion in 4.2 ) ,
the use of cde instead of cde2 for the small amorphous silicates slightly increased the amorphous pyroxene to amorphous olivine ratio .
also , 2 @xmath5 m radius solid grains of amorphous olivine were used instead of 5 @xmath5 m radius porous amorphous olivine grains ; however , as noted before , the opacity profile of 5 @xmath5 m porous amorphous silicate grains is nearly identical to that of 2 @xmath0 solid amorphous silicate grains . in addition
, graphite was used in place of amorphous carbon to give opacity continuum .
gg tau a has the next most complex 10@xmath0 feature .
gg tau a is binary , and @xcite estimate masses from @xcite stellar evolutionary models of [email protected]@xmath2 for aa and [email protected] @xmath2 for ab .
the separation of the two is 0@xmath1025 ( @xmath4 35au ) .
our spectrum is the sum of gg tau aa and gg tau ab . similar to fm tau and ip tau , the 58@xmath0 continuum from gg tau a is well in excess of the stellar photospheres , which led @xcite to conclude that there is a full inner disk in gg tau a. the circumbinary dust disk outside of 35au from the center of mass of the gg tau a pair would be too cold to give rise to the 10@xmath0 emission seen from gg tau a. grains of composition of `` astronomical silicates '' from @xcite at 252k in radiative equilibrium with a blackbody of radius 2@xmath65 and @xmath66=4000k ( representing the radiation from one star of the gg tau a pair ) would be located @xmath61au from the star .
the dust temperature of 252k ( table 2 ) implies that silicate grains are located @xmath61au from either aa or ab ( or both ) .
this is consistent with the finding by @xcite of co emission originating less than 2au from one ( or both ) of the two stars of the gg tau a system .
following @xcite , regarding disk partitioning in multiple systems , it is assumed that the ir excess must originate from a circumstellar disk around aa or ab , or both .
white et al . ( 1999 ) report h@xmath13 emission of 57@xmath67 equivalent width ( ew ) from aa but only 16@xmath67 ew from ab , suggesting a more substantial inner disk around the aa component
. however , this disk must be truncated outward of a few au , otherwise the interactions with the ab component would disrupt a larger disk .
the emissivity peaks at 9.5@xmath0 , indicating large amounts of amorphous pyroxene .
mass fractions of large silicate grains similar to tw hya are indicated , definitely higher than those for fm tau and ip tau .
a prominent 11.3@xmath0 feature indicates a forsterite mass fraction similar to that for ip tau .
quartz is suggested by possible features at 12.4 and 12.8@xmath0 . the irs mapping - mode observation of gg tau a also included signal from gg tau b in the sl slit .
gg tau b is about 10@xmath68 due south of gg tau a , and it is also a binary @xcite .
the position angle of the sl slit for each position of the gg tau a map was @xmath6344@xmath69 , close to the position angle of the separation of gg tau a and gg tau b , so gg tau a and b were close to maximally separated in the short - low slits of two of the map positions .
the spectrum of gg tau b was extracted from the two map positions best centered on its position .
the short - low slit is 2 pixels ( 3@xmath106 ) wide , and the extraction aperture used varies from as little as 3.3 pixels ( 6 ) at the shortest wavelength of first order to as much as 5 pixels ( 9 ) at the longest wavelength .
this extraction region was centered on gg tau b ; at the longest wavelengths , the edge of the extraction region in sl closest to gg tau a is @xmath623 pixels away from the center of the psf of gg tau a , which is believed to be sufficiently far away to minimize contribution of signal from gg tau a. white et al .
( 1999 ) report the two components of the gg tau b binary , ba and bb , to be separated by 1@xmath1048 ( 207au ) , which is smaller than the short - low beam .
the spectrum shown for gg tau b therefore includes signal from both ba and bb , as well as from circumstellar disk(s ) around either or both stars .
white et al . ( 1999 ) assign masses of 0.12 @xmath11 0.02@xmath2 for ba and [email protected]@xmath2 for bb , and report h@xmath13 ew of @xmath620@xmath67 for ba and 2043@xmath67 for bb .
this puts both ba and bb near the hydrogen - burning mass limit .
the 10@xmath0 feature from the gg tau b pair indicates a higher mass fraction of crystalline grains than any of coku tau/4 , tw hya , fm tau , ip tau , or gg tau a. because the 10@xmath0 emission from gg tau b is seen and also because the separation between ba and bb is @xmath6200au , it is believed that a circumstellar disk(s ) exist(s ) around either ba or bb , or both .
it is unclear to which component , ba or bb ( or both ) , the gg tau b disk emission belongs .
currently , we have no long - wavelength data for gg tau b , so instead of deriving a silicate dust temperature for gg tau b , the same dust temperature for gg tau a , 252k , is used .
the peak of the 10@xmath0 feature is much flatter than that of any of the previous sources , with sharp inflections at 9.4 and 11.2@xmath0 defining the plateau . due to the low signal ( @xmath680mjy at 10@xmath0 ) , the s
/ n is rather low , so only a rough fit to the 10@xmath0 feature was attempted .
the narrow peak in emissivity at @xmath49.4@xmath0 suggests crystalline pyroxene .
the plateau shape of the rest of the 10@xmath0 feature is due to larger grains and an admixture of small crystalline grains of other composition forsterite , in this model .
the emissivity and dust model deviate longward of 12.5@xmath0 .
noting that the flux at sl wavelengths for gg tau a is roughly 10 times that for gg tau b , and noting that the extraction aperture is largest for the longest wavelengths of the first order of sl , we attribute the @xmath2012.5 @xmath5 m rise of emissivity above model to contamination of signal from gg tau a. hen 3 - 600 is another multiple system with mid - ir emission .
@xcite resolve the hen 3 - 600 system into a and b components and measure a separation of components of 1@xmath104 ( 70au assuming the hen 3 - 600 pair belongs to the tw hydrae association ) .
in addition , hen 3 - 600 a itself is a spectroscopic binary @xcite .
none of the components of the hen 3 - 600 system are resolved in any of the instrument slits @xcite , but jayawardhana et al .
( 1999 ) determined that the circumstellar disk is associated with the a pair .
due to the unknown separation of components of the spectroscopic binary , the disk(s ) in this system can not be assigned to either of the components .
it is assumed that the a pair are separated by less than 1 au and that the irs spectrum for hen 3 - 600 a arises from a circumbinary disk around the a pair .
the combined luminosity of the two components from @xcite are listed in table 1 .
@xcite and @xcite report that the 10@xmath0 feature for this object indicates amounts of crystalline silicates comparable to those of amorphous silicates .
the derived dust temperature for hen 3 - 600 a is @xmath6229k , higher than for other transitional disks .
this means that this object has more grains in its inner disk than other transitional disks , consistent with a transition radius of 1.3au ( @xcite ) , the closest ( to the star ) among all transitional disks .
the 10@xmath0 feature ( figure 6 ) shows a 911@xmath0 plateau , along with a strong , narrow feature at 9.2@xmath0 .
the 9.2@xmath0 feature indicates quartz in the cde distribution @xcite .
silica was previously identified by @xcite and @xcite .
alpha quartz produces smaller peaks at 12.4 and 12.8@xmath0 , which may correspond to a small feature centered at 12.6@xmath0 in the emissivity .
alternatively , other polymorphs and forms of silica give single - peaked features at @xmath612.6@xmath0 @xcite , so it is quite possible that other forms or polymorphs of silica may be present in hen 3 - 600 a. in addition to quartz , a larger mass fraction of the material responsible for the continuum opacity than for any other objects in this sample is needed to fit the 1214@xmath0 continuum .
small amorphous carbon grains are assumed , but it could just as well come from very large silicate grains .
the overall flatness of the 10@xmath0 plateau indicates a small amount of crystalline pyroxene .
forsterite is indicated by the sharp 11.3@xmath0 edge of the plateau , and small and large grains of amorphous pyroxene are added to smooth the crystalline silicate features between 9.2 and 11.3@xmath0 .
the emissivity rises above the dust model from 13 to 14@xmath0 , although much less drastically than for gg tau b. such a feature in the emissivity could be due to a dust component other than what we use for hen 3 - 600 a. the derived emissivity of fn tau also indicates a large crystalline silicate mass fraction .
the plateau top to the 10@xmath0 feature is not flat , but it slopes downward to longer wavelengths , moreso than for hen 3 - 600 a. the derived emissivity of fn tau peaks at 9.3@xmath0 , a bit longer than that of hen 3 - 600 a. it also has minor peaks at 9.9 , 10.6 , 11.2 , and 11.5@xmath0 .
all of these peak positions are highly indicative of crystalline pyroxene .
amorphous pyroxene smooths the narrow subfeatures at the top of the 10@xmath0 sloping plateau , while simultaneously accentuating the 9.3@xmath0 feature .
approximately 1% by mass of quartz is indicated by the model . like tw hya
, there is a `` shoulder '' at wavelengths shorter than 8.7@xmath0 that may indicate a silica type other than @xmath13 quartz . like gg tau b and hen 3 - 600 a
, the dust model underestimates the emissivity at wavelengths near 14@xmath0 ( see the discussion of gg tau b ) .
the v410 anon 13 emissivity is similar to those of fn tau and hen 3 - 600a .
v410 anon 13 is a very low - mass star ( @xmath60.1@xmath2 ; * ? ? ?
* ) of spectral type m5.75 ; quite remarkably , it is surrounded by a full , flared accretion disk . like fn tau , the derived emissivity peaks at 9.3@xmath0 , indicating crystalline pyroxene as opposed to quartz .
no clear identification can be made for any features ( or lack thereof ) between 12 and 13@xmath0 due to relatively high spectral noise , so a small amount of quartz is not ruled out . like hen 3 - 600
a , the rest of the 10@xmath0 feature is a fairly featureless flat plateau with a sharp knee at 11.2@xmath0 .
the sharp 11.2@xmath0 knee indicates forsterite .
small grains of amorphous pyroxene emphasize the 9.3@xmath0 feature and dampen the crystalline features of forsterite and crystalline pyroxene along the top of the 10@xmath0 plateau . as with fn tau , the 8.7@xmath0 shoulder and the emissivity between 13 and 14@xmath0
are both underestimated by the dust model . because the extinction correction was only applied for v410 anon 13 , the effects of changing the extinction correction on the dust model for this object were explored .
as discussed previously , it is believed that any local extinction would be very similar to the interstellar extinction , so only the amount of extinction was varied .
when the optical depth of intervening material was doubled ( from @xmath19=0.32 to @xmath19=0.64 ) , the derived dust temperature increased from 256 to 289k .
the inferred mass fraction of amorphous pyroxene increased from 48% to 61% , quartz decreased from 2% to 1% , crystalline pyroxene decreased from 11% to 10% , forsterite decreased from 5% to 2% , and amorphous carbon decreased from 34% to 26% .
the @xmath64/dof increased slightly from 1.9 to 2.1 , but the fit still looked reasonable .
the increase of derived dust temperature results from the opacity of interstellar silicates being higher in the 10@xmath0 feature than in the 20@xmath0 feature .
the modest increase in the mass of amorphous pyroxene is attributed to the variation of the amount of extinction correction over the 10@xmath0 feature .
the extinction correction is greatest near @xmath69.7@xmath0 , vertically `` stretching '' the 10@xmath0 feature so that more amorphous pyroxene , peaking at 9.4@xmath0 , is needed .
the extinction correction also has the effect of `` rounding off '' any narrow peaks due to crystalline silicates , such as the 11.3@xmath0 peak best fit by forsterite , thus decreasing the amount of crystalline silicates .
in addition , this stretching raises the contrast of the 10@xmath0 feature above the @xmath2013@xmath0 continuum , effectively decreasing the amount of continuum opacity required from amorphous carbon .
the most troublesome emissivity to characterize was that of cy tau .
while the 58@xmath0 excess of this object is similar to those of the other full disks in our sample , the contrast of the 10@xmath0 excess to the underlying continuum is the lowest in the entire sample ( very low @xmath41 ; see table 2 ) .
for this reason , the derived emissivity is poorly determined over the entire 10@xmath0 feature .
what can be discerned is a fairly narrow feature around 9.3@xmath0 ( as with fn tau and v410 anon 13 ) , and a noisy downward slope to @xmath611.4@xmath0 , at which point the feature sharply drops . while large grains can provide flat - topped 10@xmath0 features of low contrast
, they can not produce the 9.3@xmath0 opacity sharp peak ; therefore , only a modest amount of large grains are included . between the local minima at 11.8 and 13.2@xmath0 in the emissivity , a small feature rises above the pixel - to - pixel noise , followed by a rise past 13.5@xmath0 to longer wavelengths .
crystalline pyroxene can account for the 9.3@xmath0 peak , while the 11.4@xmath0 inflection indicates forsterite .
as with v410 anon 13 , quartz is consistent within the noise near 12.5@xmath0 .
in addition , quartz provides , as with hen 3 - 600a , a boost to the crystalline pyroxene at 9.3@xmath0 to increase its prominence , while at the same time giving rise to spectral features in the 12 to 13@xmath0 region accounting for some of the noisy features between 11.8 and 13.2@xmath0 .
amorphous pyroxene is included to provide overall roundness to the 10@xmath0 feature . as with fn tau
, the dust model underestimates the derived cy tau emissivity near 14@xmath0 .
@xcite conclude that no more than 5% of the si in the diffuse ism is in crystalline silicates .
@xcite place an upper limit of @xmath62.2% for the silicate dust toward the galactic center .
the result of @xmath60.1% mass fraction of crystalline silicates ( forsterite , crystalline pyroxene , and quartz ) for coku tau/4 and nearly zero crystalline silicates for dm tau and gm aur ( table 4 ) , is therefore consistent with nearly no processing of silicate dust in these disks , and with the material in these disks having originated from the ism .
this seems to indicate that little processing occurs in the outer disks , or , if processing does occur in the inner disk , the dust in the inner disk is not transported efficiently to the outer disk .
perhaps the material in the inner optically thin disk of gm aur was located between 10 and 24au immediately before the planet formed near 24au .
this material would have experienced no processing , as in the case of coku tau/4 .
perhaps a planet opened up a disk gap at @xmath624au , halting inward accretion at 24au of material from the outer disk .
the disk inside of 24au would continue accreting , and perhaps the last of this inner disk of nonprocessed dust is seen just before it gets accreted onto the star . in figure 7
the mass fraction of crystalline silicates @xmath70/@xmath71 versus wall radius is plotted for the five transitional disks .
note the lack of crystallinity for disks with walls beyond 2au .
the spectra of the rest of the 13myr old taurus sample , all of which have inner disks more substantial than those of coku tau/4 , dm tau , and gm aur , indicate that crystalline silicates constitute at least @xmath60.5% ( by mass ) of all silicates in each disk .
whatever is responsible for producing the crystalline grains in t tauri disks therefore appears linked to the inner 3au disk region , a finding in line with the spectrointerferometric observations of herbig ae / be disks @xcite .
a substantial crystalline component is in place in the `` full '' t tauri disks ( defined in 3.2 ) at 13myr . in our sample of ttss
, a majority of the silicate dust mass is always found in amorphous silicates .
the disks in the sample with lower mass fractions of crystalline silicates indicate forsterite but no crystalline pyroxene , while the disks with higher crystalline silicate mass fractions indicate both types of crystalline silicates .
this is consistent with the trend reported by @xcite regarding the disks around herbig ae stars . confirming this trend will require a larger sample and a more thorough investigation of the degeneracy between forsterite and crystalline pyroxene ( see discussion in 3.6 ) .
all spectra except v410 anon 13 indicate some large grains , but a greater crystalline mass fraction does not appear to correlate with a greater large - grain mass fraction ( see figure 8) for the t tauri objects in our sample .
see the discussion in 3.5 on how grain sizes and shapes assumed in this study compare to those in other studies @xcite . because quartz accounts for less than @xmath63% of the total dust mass in the models , this implies that the mass fraction of forsterite and crystalline pyroxene does not correlate with that of the 5@xmath0 sized porous amorphous silicate grains .
it seems reasonable that , over time , the mass fraction of crystalline grains would increase , as should the mass fraction of large grains in disks around stars .
however , crystallization and grain growth are not the same processes crystallization involves changing the grain structure at the atomic level , while grain growth involves the gradual clumping together of small grains to aggregate into large grains .
the rates of increase of the crystalline grain mass fraction and of the large grain mass fraction should not necessarily be the same for even one single yso disk .
furthermore , silicate grain growth and crystallization of silicate dust might proceed at different rates in disks with different accretion rates , mass , etc .
it is not clear why large - grain mass fraction and crystalline grain mass fraction should be correlated in herbig ae / be stars @xcite but not in t tauri stars .
it is also curious that v410 anon 13 indicates a substantial mass fraction of crystalline silicates but a negligible mass fraction of large grains .
if , as proposed by @xcite , large grains in disks settle more quickly than smaller grains toward the optically thick midplane , and vertical convection is negligible ( as is expected due to the vertical temperature inversion in disk atmospheres ) , then the lack of larger grains in a disk may simply be due to the fact that large grains have settled from the optically thin atmosphere of the disk , where they can be seen , down to the optically thick midplane
. a high mass fraction of very large grains could be present in the v410 anon 13 disk atmosphere and only produce a featureless continuum .
report that herbig ae / be stars of mass greater than 2.5@xmath2 have consistently high mass fractions of crystalline silicates ( @xmath420% to @xmath430% ) , and that herbig ae / be stars less massive than 2.5@xmath2 have crystalline silicate mass fractions varying widely between 0 and @xmath620% .
it is found that , like the lower mass herbig ae / be stars , ttss of mass @xmath61.2@xmath2 down to @xmath60.1 @xmath2 exhibit large dispersions in the mass fractions of crystalline silicates .
silicates in the disks around the lowest mass stars in our sample fn tau , v410 anon 13 , gg tau b have the highest crystalline percentages ( between @xmath627% and @xmath630% ) in our sample .
according to the mass tracks from several different stellar evolutionary models , fn tau ( spectral type m5 ; * ? ? ?
* ) is less massive , ( @xmath60.28@xmath2 ; * ? ? ?
* ) than all other stars in our sample except v410 anon 13 and the component stars of the gg tau b pair .
the disks in our sample with the highest mass fractions of crystalline silicates surround the lowest mass stars .
a study of silicate crystallinity as a function of stellar mass for a much larger sample is underway ( d. m. watson et al .
2006 , in preparation ) .
the mass fraction of crystalline silicates @xmath70/@xmath71 versus stellar mass is plotted in figure 9 . using the mass fraction of only forsterite and crystalline pyroxene instead of @xmath70 in figure 9
would not greatly change it as quartz is never a major component of the total crystalline grain mass .
the emissivities of hen 3 - 600 a and gg tau b ( both known as binary systems ) , indicate some of the highest mass fractions of crystalline grains in our sample ; the emissivity of the binary gg tau a ( which is coeval with gg tau b , according to @xcite ) , however , indicates a low mass fraction of crystalline silicates .
numerous mechanisms have been proposed for processing amorphous silicate grains into crystalline silicates .
based on the midi ( mid - ir interferometric instrument ) spectra , @xcite argue that in the inner disk regions ( @xmath12au ) , crystalline silicate grains are condensed directly from the gas phase .
alternatively , solid amorphous silicates may be thermally annealed into crystalline grains ; in thermal annealing , grains are heated sufficiently that constituent atoms redistribute themselves in the grain in more energetically favored positions , forming a regular crystalline lattice @xcite .
various mechanisms have been proposed to transport grains radially from the warm inner disk regions , where crystalline silicates may be produced , to the outer disk regions .
these include convection and turbulent mixing @xcite , turbulent diffusion and large - scale circulation currents @xcite , and _
x_-winds @xcite .
spectra of coku tau/4 and dm tau indicate that almost no crystalline silicates exist in the regions emitting the mid - ir spectra . for coku
tau/4 , this region is mostly the vertical wall at @xmath610au . for dm tau , the region giving rise to much of its mid - ir spectrum is a combination of the wall at @xmath63au and the cooler optically thin atmosphere immediately beyond the wall @xcite .
the lack of crystalline silicate grains in the outer disks of these two objects ( beyond @xmath63au for dm tau and beyond @xmath610au for coku tau/4 ) seems to suggest that very few crystalline silicate grains were radially transported to their outer disks ; however , this may be expected in the case of planet formation and inner disk clearing . after a planet forms , it can clear out a gap in a disk , and any newly formed crystalline silicate dust interior to the planet s orbit will either be accreted onto the planet or onto the star . furthermore
, crystalline dust in the upper layers of the outer disk will accrete inward onto the newly formed planet . local production of crystalline grains is an alternative to radial transport .
lightning has been proposed @xcite to heat dust grains and chondrules , but it is not clear exactly whether lightning in yso disks might be generated @xcite . @xcite propose _ in situ _ annealing at disk radii between @xmath65 and 10au , summarizing evidence that chondrules ( which contain crystalline silicates ) formed from circumstellar material in young disks between @xmath65 and 10au ; therefore , they look to that region as the site of dust grain annealing .
harker & desch ( 2002 ) propose that submicron- to micron - sized dust grains are heated and annealed by thermal exchange with surrounding gas , itself heated by a disk shock front ; the rate of cooling of the grains is then determined by the rate of cooling of the surrounding disk gas .
their simulations , using models described by desch & connolly ( 2002 ) , indicate that silicate grains can be annealed in 5kms@xmath38 shocks out to 10au . in some simulations , 250@xmath0 radius chondrules were completely evaporated . according to @xcite
, shock fronts have been seen in simulations of gravitationally unstable disks as transient phenomena . to initiate gravitational instability
, @xcite asserts that the ratio of the disk mass to the stellar mass must be around 0.3 to 0.5 ; however , in some numerical simulations , disk - to - star mass ratios as low as 0.05 @xcite initiate the growth of gravitational instability in a disk , resulting in the formation of a spiral arm .
this spiral arm would then be the site of shock annealing . in an attempt to characterize the degree of likelihood of such shock processing to occur , disk mass relative to stellar mass in our small sample
is considered .
it is not known to which component of the gg tau a binary the gg tau a disk emission belongs , so it is not included in our discussion of the effect of the ratio of the disk mass to the stellar mass on the mass fraction of crystalline grains ; for the same reason , gg tau b is not included .
in addition , there are no disk mass estimates for v410 anon 13 and hen 3 - 600 a , so they are not included .
in figure 10 we plot the disk - to - star mass ratio versus the mass fraction of crystalline silicates .
there is no clear correlation between the disk - to - star mass ratio and the silicate crystallinity .
as with figures 8 and 9 , when quartz is excluded from the total crystalline mass fraction , the conclusion does not change ( i.e. there is no correlation between the disk - to - star mass ratio and the total mass of crystalline pyroxene and forsterite ) . it is possible that planet formation may obscure any correlation between the disk - to - star mass ratio and the crystalline silicate mass fraction . gravitational instabilities may induce disk shocks .
if planets form by gravitational instability , which should be correlated with the disk - to - star mass ratio , then planets should form when disk shocks are in the process of annealing silicate grains .
any crystalline silicates produced by disk shocks , which are interior to the forming planet , would then either accrete onto the planet or onto the star , thus eventually removing any evidence of them . the relation ( or lack thereof ) of the disk mass - to - star mass ratio with the crystalline mass fraction for a much larger sample of objects in the taurus - auriga star - forming region
will be studied in a forthcoming paper . in figure 11
is plotted for the full disks the mass fraction of crystalline silicates versus @xmath41 , the measure of the contrast of the 10@xmath0 feature to the underlying continuum .
note that the disks with lower crystallinity have higher @xmath41 , and those with high degrees of crystallinity have lower @xmath41 .
the mass fraction of large silicate grains versus @xmath41 is plotted in figure 12 . here
an upper envelope , decreasing for increasing @xmath41 , to the large grain fraction may be present .
others have observed in herbig ae / be stars @xcite and t tauri stars @xcite a trend of decreasing the 10@xmath0 feature contrast with increased dust processing ( grain growth and/or dust crystallization ) .
the decrease of both the crystallinity and the large grains mass fractions with increasing @xmath41 in our sample of 12 t tauri disks is consistent with the earlier studies . at 10myr ,
hen 3 - 600 a presents a disk with a large mass fraction of crystalline silicates , while tw hya does not .
tw hya ( @xmath60.7@xmath2 ; see webb et al .
1999 and siess et al .
2000 ) is hypothesized to have a protoplanetary or planetary companion orbiting between @xmath62 and 3au @xcite and is reported to have a fairly massive disk of @xmath40.03 @xmath2 @xcite .
@xcite show that the optically thin disk inside of @xmath63.3au is responsible for most of the 10@xmath0 emission in tw hya .
while the spectrum of tw hya does indicate a small amount of forsterite and quartz , and a substantial fraction of large grains , it does not indicate crystalline pyroxene ; the derived crystalline mass fractions for tw hya are similar to those of gg tau a and ip tau , but tw hya is much older than those stars .
perhaps the protoplanetary companion to tw hya orbiting at 2 - 3au prevents outward radial transfer of crystalline silicate grains produced in the inner disk ; alternatively , perhaps the small star - planet separation allows only a small zone in which shocks can form , annealing fewer silicate grains .
the transition from optically thin inner disk to optically thick outer disk is modeled to be at @xmath61.3au in hen 3 - 600 a @xcite , indicating a more substantial disk in hen 3 - 600 a than in tw hya .
hen 3 - 600 a is the only one transitional disk with a substantial degree of crystallinity ( see figure 7 ) .
the denser disk between 1.3 and 3.3au in hen 3 - 600 a might lead to either more inner disk crystalline silicate production and freer radial transport of crystalline grains or more substantial shock processing than in tw hya .
there is an indication of increasing mass fraction of quartz with decreasing mass fraction ratio of amorphous olivine to amorphous pyroxene ( figure 13 ) .
the irs spectra of tw hya , hen 3 - 600 a , and gg tau a all indicate the presence of quartz and low amorphous olivine to amorphous pyroxene mass ratios .
we caution , however , that our sample size is small ( 12 spectra ) , and that further study of this possible trend , using a much larger sample of spectra , will be needed to verify this finding .
furthermore , some of the mass in quartz in these disks could be in forms of silica other than @xmath13 quartz , which would affect this trend .
these issues will be explored in a future paper .
the 10@xmath0 features in 10 tts spectra from the taurus - auriga star - forming region and 2 spectra from the tw hydrae association are modeled by deriving the emissivities from irs spectra of the optically thin dust emission . through an iterative process of varying dust grain mass fractions of various dust species to match the dust cross - sections to the derived emissivities and using the derived @xmath39(@xmath420 @xmath5m)/@xmath39(@xmath410 @xmath5 m ) ratios to recompute the dust temperature ( and therefore self - consistently define the emissivity ) , the modeled emissivities are fit with a multi - component dust model . to test the simple single - temperature approach of deriving and fitting the 10@xmath0 feature emissivities , a wavelength - dependent dust cross - section similar to the one generated from the fit to ip tau was used to determine the emergent spectrum in a model of its emission , and it was found that the fit to the spectrum of ip tau requires a significant source of absorption at visible / near - ir wavelengths . to fit the interstellar extinction
, @xcite also required a source of absorption at these wavelengths in addition to that from known silicates .
consequently , graphite and `` astronomical silicates '' were used to determine the radial and vertical temperature structure of the model disk for ip tau , and a dust mixture similar to the best - fit dust mixture for this object was used to compute the emergent spectrum from the model disk .
this full radiative transfer model gave an excellent fit to the 512@xmath0 spectrum of ip tau , supporting the simplified emissivity modeling technique used here .
the model predicts too much emission at wavelengths longer than 12@xmath0 , indicating too much emission in the model from cooler dust grains , suggesting there is settling of dust from the disk atmosphere to the midplane .
our 12 object sample indicates the following : * almost all the t tauri disks show some degree of silicate crystallinity , irrespective of star mass or age .
* transitional disks , disks whose inner portions are either partially or nearly totally cleared of small dust grains , usually indicate very few crystalline silicate grains .
* theories of radial transport of crystalline silicates in t tauri disks are constrained to explain the ism - like lack of crystalline silicates outward of 3au in dm tau and outward of 10au in coku tau/4 , both of which are 13 myr old systems . *
no obvious correlation of the mass fraction of crystalline silicate grains with that of large silicate grains exists for the t tauri sample .
this contrasts with the results of @xcite for herbig ae / be systems ( see discussion in 3.5 regarding differences in grain shape and size between this study and bouwman et al . [
2001 ] and van boekel et al .
[ 2005 ] ) . *
the 10@xmath0 features always indicate a majority of mass in amorphous silicate grains . *
crystalline pyroxene is usually accompanied by forsterite , but the reverse is not necessarily true .
* very low mass stars can have relatively large amounts of crystalline silicates in their surrounding disks . *
no clear trend exists between the mass fraction of crystalline silicates and the disk - to - star mass ratio . * for full disks , high contrast of the 10@xmath0 feature , as measured by @xmath41 , indicates both low crystallinity and small mass fraction of large grains ; decreasing contrast of the 10@xmath0 feature for these disks indicates increasing crystallinity and a range of mass fractions of large grains .
this is consistent with studies of herbig ae / be stars @xcite and previous studies of t tauri stars @xcite .
* there is an indication that higher quartz mass fraction accompanies lower amorphous olivine to amorphous pyroxene ratio .
one of the surprising results of this study is the lack of correlation of the silicate crystallinity with anything else .
this may reflect the diversity in the evolutionary histories of disks ( e.g. , some make more crystalline silicates than others ) , which may be a consequence of initial conditions ( e.g. , initial disk mass and angular momentum ) and/or environment ( e.g. , binarity ) .
considerable diversity may result from the process of giant planet formation .
for example , suppose that within the first 13myr of existence , all disks accumulate crystalline silicate grains in their inner disks either by inner disk heating and outward radial transport or by local annealing of silicate grains .
suppose further that a giant planet forms in the inner disk . according to @xcite , the planet clears a gap in the disk and absorbs the accretion from the outer disk .
the inner disk is cleared by accretion onto the star in @xmath610@xmath72 yr , leaving a transitional disk with the characteristics of the disk around either coku tau/4 or dm tau .
the remnant outer disk has very few crystalline silicate dust grains remaining , also like coku tau/4 .
if the disk is sufficiently massive , as is the case with dm tau , the planet can migrate radially inward . if this planet formation and migration scenario happens with sufficient frequency in t tauri disks , a large number of t tauri disks with near ir excesses not indicating transitional disks but with 10@xmath0 features indicating low crystallinity , like fm tau , will be observed .
both giant planet formation and crystallization can depend on the disk / star mass ratio , but crystallized dust can be removed after the planet forms . in a future paper , the relative abundances of various dust species in a larger sample of t tauri disks will be quantified , and the abundances of dust species will be compared to other dust species and also to various stellar properties . another goal is to incorporate the best - fit dust cross - sections for our t tauri objects into the sophisticated radiative transfer models described by @xcite , and @xcite to model these objects seds and spectra self - consistently over much wider wavelength ranges .
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 nasa contract 1407 .
support for this work was provided by nasa through contract 1257184 issued by jpl / caltech and through the spitzer fellowship program , under award 011 808 - 001 , and jpl contract 960803 to cornell university , and cornell subcontracts 31419 - 5714 to the university of rochester .
the authors thank ciska markwick - kemper for offering 10@xmath0 data for gcs3 and for helpful comments on the paper .
the authors also acknowledge helpful comments from luke keller .
p. d. acknowledges grants from papiit , dgapa , unam and conacyt , mxico .
n. c. and l. h. acknowledge support from nasa grant nag5 - 13210 , stsci grant ar-09524.01-a , and nasa origins grant nag5 - 9670 .
a.l . acknowledges support from the university of missouri summer research fellowship , the university of missouri research board , and the nasa award p20436 .
smart was developed by the irs team at cornell university and is available through the _ spitzer _
science center at caltech .
this publication makes use of the jena - st .
petersburg database of optical constants @xcite .
in our attempts to fit the dust emission feature at @xmath611.3@xmath0 in many of our emissivities , we unexpectedly found a very close resemblance between the opacity profile of forsterite grains with a cde shape distribution and that of porous forsterite grains with a volume fraction of vacuum @xmath60=0.6 ( i.e. 60% of the grain volume is vacuum ) and a fractal dimension of @xmath54=2.766 .
for these porous grains , we used the bruggeman effective medium theory ( bohren & huffman 1983 ) and the dielectric constants of forsterite of @xcite to generate the effective dielectric constants for each crystalline axis . for a given crystalline axis ,
we generate opacity from mie theory using the effective dielectric constants for that axis , assuming a spherical grain radius of 0.1@xmath0 .
we then obtain the opacity for the porous forsterite grain by averaging the opacities of all three crystalline axes .
the resulting porous forsterite opacity is nearly identical to that derived using the same set of dielectric constants and assuming a cde shape distribution ( figure 14 ) .
we made the same comparison using parameters for the ordinary and extraordinary rays of @xmath13 quartz ( wenrich & christensen 1996 ) .
as for forsterite , we used @xmath60=0.6 , @xmath54=2.766 , and grain radius of 0.1@xmath0 ; however , for quartz we add the sum of of the opacity generated from the ordinary ray and of that from the extraordinary ray .
as for forsterite , the quartz opacities generated assuming porous grains and grains with a cde distribution are nearly identical ( see figure 14 ) .
this suggests that the near - equivalence of the bruggeman emt plus mie theory to cde is mathematical in nature and not specific to the optical properties of the dust material in question .
adams , f. c. , lada , c. j. , & shu , f. h. 1987 , , 312 , 788 andrews , s. m. , & williams , j. p. 2005
, , 631 , 1134 baraffe , i. , chabrier , g. , allard , f. , & hauschildt , p. h. 1998 , , 337 , 403 beckwith , s. v. w. , sargent , a. i. , chini , r. s. , & guesten , r. 1990 , , 99 , 924 bohren , c. f. , & huffman , d. r. 1983 , absorption and scattering of light by small particles , new york : wiley boss , a. p. 2004
, , 616 , 1265 boss , a. p. , & durisen , r. h. 2005a , , 621 , l137 boss , a. p. , & durisen , r. h. 2005b , asp conf .
ser . 341 : chondrites and the protoplanetary disk , 341 , 821 bouwman , j. , meeus , g. , de koter , a. , hony , s. , dominik , c. , & waters , l. b. f. m. 2001 , , 375 , 950 bradley , j. 2003 , lnp vol . 609 : astromineralogy , 609 , 217 brott , i. , & hauschildt , p. h. 2005 , esa sp576 : the three - dimensional universe with gaia , 565 brucato , j. r. , colangeli , l. , mennella , v. , palumbo , p. , & bussoletti , e. 1999 , , 348 , 1012 calvet , n. , patino , a. , magris , g. c. , & dalessio , p. 1991
, , 380 , 617 calvet , n. , magris , g. c. , patino , a. , & dalessio , p. 1992 , rev . mex .
astrof . , 24 , 27 calvet , n. , dalessio , p. , hartmann , l. , wilner , d. , walsh , a. , & sitko , m. 2002 , , 568 , 1008 calvet , n. , et al .
2005 , , 630 , l185 chihara , h. , koike , c. , tsuchiyama , a. , tachibana , s. , & sakamoto , d. 2002 , , 391 , 267 cohen , m. 1973 , , 164 , 395 cohen , m. , & witteborn , f. c. 1985 , , 294 , 345 cohen , m. , megeath , s. t. , hammersley , p. l. , martn - luis , f. , & stauffer , j. 2003 , , 125 , 2645 dalessio , p. , canto , j. , calvet , n. , & lizano , s. 1998 , , 500 , 411 dalessio , p. , calvet , n. , hartmann , l. , lizano , s. , & ca nt ' o , j. 1999 , , 527 , 893 dalessio , p. , calvet , n. , & hartmann , l. 2001 , , 553 , 321 dalessio , p. , et al .
2005 , , 621 , 461 dalessio , p. , et al .
2006 , , 638 , 314 desch , s. j. , & connolly , h. c. 2002 , meteoritics & planet .
, 37 , 183 dorschner , j. , begemann , b. , henning , t. , jaeger , c. , & mutschke , h. 1995 , , 300 , 503 dorschner , j. 2003 , lnp vol . 609 : astromineralogy , 609 , 1 draine , b. t. , & lee , h. m. 1984 , , 285 , 89 draine , b. t. 2003 , , 41 , 241 dyck , h. m. , capps , r. w. , forrest , w. j. , & gillett , f. c. 1973 , , 183 , l99 etchepare , j. , merian , m. , & kaplan , p. 1978 , , 68 , 1531 fabian , d. , jger , c. , henning , t. , dorschner , j. , & mutschke , h. 2000 , , 364 , 282 fabian , d. , henning , t. , j " ager , c. , mutschke , h. , dorschner , j. , & wehrhan , o. 2001 , , 378 , 228 forrest , w. j. , houck , j. r. , & reed , r. a. 1976 , , 208 , l133 forrest , w. j. , & soifer , b. t. 1976 , , 208 , l129 forrest , w. j. , mccarthy , j. f. , & houck , j. r. 1979 , , 233 , 611 forrest , w. j. , et al . 2004 , , 154 , 443 furlan , e. , et al . 2005a , , 621 , l129 furlan , e. , et al . 2005b , , 628 , l65 gail , h
2004 , , 413 , 571 geoffray , h. , & monin , j .- l . 2001 , , 369 , 239 gillett , f. c. , low , f. j. , & stein , w. a. 1968 , , 154 , 677 harker , d. e. , & desch , s. j. 2002 , , 565 , l109 harker , d. e. , wooden , d. h. , woodward , c. e. , & lisse , c. m. 2002 , , 580 , 579 hartmann , l. , calvet , n. , gullbring , e. , & dalessio , p. 1998
, , 495 , 385 henning , t. , ilin , v. b. , krivova , n. a. , michel , b. , & voshchinnikov , n. v. 1999 , , 136 , 405 herczeg , g. j. , wood , b. e. , linsky , j. l. , valenti , j. a. , & johns - krull , c. m. 2004 , , 607 , 369 higdon , s. j. u. , et al .
2004 , , 116 , 975 honda , m. , kataza , h. , okamoto , y. k. , miyata , t. , yamashita , t. , sako , s. , takubo , s. , & onaka , t. 2003 , , 585 , l59 houck , j. r. , et al .
2004 , , 154 , 18 jayawardhana , r. , hartmann , l. , fazio , g. , fisher , r. s. , telesco , c. m. , & pi~ na , r. k. 1999 , , 520 , l41 jones , t. w. , & merrill , k. m. 1976 , , 209 , 509 kemper , f. , vriend , w. j. , & tielens , a. g. g. m. 2004 , , 609 , 826 ( erratum : 633 , 534 ) kenyon , s. j. , & hartmann , l. 1987 , , 323 , 714 kenyon , s. j. , & hartmann , l. 1995 , , 101 , 117 kessler , m. f. , et al . 1996 , , 315 , l27 kessler - silacci , j. e. , hillenbrand , l. a. , blake , g. a. , & meyer , m. r. 2005 , , 622 , 404 kessler - silacci , j. , et al .
2006 , , 639 , 275 koike , c. , shibai , h. , & tuchiyama , a. 1993 , , 264 , 65 levesque , e. m. , massey , p. , olsen , k. a. g. , plez , b. , josselin , e. , maeder , a. , & meynet , g. 2005 , , 628 , 973 li , a. , & draine , b. t. 2001 , , 550 , l213 li , a. , & greenberg , j. m. 1997 , , 323 , 566 li , a. , & greenberg , j. m. 1998 , , 498 , l83 li , a. , & lunine , j. i. 2003 , , 590 , 368 lisse , c. m. , ahearn , m. f. , hauser , m. g. , kelsall , t. , lien , d. j. , moseley , s. h. , reach , w. t. , & silverberg , r. f. 1998 , , 496 , 971 mendoza , v. e. e. 1966 , , 143 , 1010 najita , j. , carr , j. s. , & mathieu , r. d. 2003 , , 589 , 931 natta , a. , meyer , m. r. , & beckwith , s. v. w. 2000 , , 534 , 838 pilipp , w. , hartquist , t. w. , morfill , g. e. , & levy , e. h. 1998 , , 331 , 121 pollack , j. b. , hollenbach , d. , beckwith , s. , simonelli , d. p. , roush , t. , & fong , w. 1994 , , 421 , 615 przygodda , f. , van boekel , r. , brahm , p. , melnikov , s. y. , waters , l. b. f. m. , & leinert , c. 2003 , , 412 , l43 quillen , a. c. , blackman , e. g. , frank , a. , & varni ` ere , p. 2004
, , 612 , l137 rogers , c. , martin , p. g. , & crabtree , d. r. 1983 , , 272 , 175 rucinski , s. m. 1985 , , 90 , 2321 shu , f. h. , shang , h. , & lee , t. 1996 , science , 271 , 1545 siess , l. , dufour , e. , & forestini , m. 2000 , , 358 , 593 simon , m. , & prato , l. 1995 , , 450 , 824 sloan , g. c. , kraemer , k. e. , price , s. d. , & shipman , r. f. 2003 , , 147 , 379 sogawa , h. , kozasa , t. , koike , c. , & suto , h. 1999 , in proceedings of the 32nd isas lunar and planetary symposium 32 , ed . h. mizutani & m. kato ( sagamihara : inst .
of space and aeronautical science ) , 175 speck , a. k. 1998 , ph.d .
thesis , university college london torres , g. , guenther , e. w. , marschall , l. a. , neuh " auser , r. , latham , d. w. , & stefanik , r. p. 2003
, , 125 , 825 uchida , k. i. , et al .
2004 , , 154 , 439 van boekel , r. , waters , l. b. f. m. , dominik , c. , bouwman , j. , de koter , a. , dullemond , c. p. , & paresce , f. 2003 , , 400 , l21 van boekel , r. , et al .
2004 , , 432 , 479 van boekel , r. , min , m. , waters , l. b. f. m. , de koter , a. , dominik , c. , van den ancker , m. e. , & bouwman , j. 2005 , , 437 , 189 watson , d. m. , et al .
2004 , , 154 , 391 watson , d. m. , et al .
2006 , in preparation webb , r. a. , zuckerman , b. , platais , i. , patience , j. , white , r. j. , schwartz , m. j. , & mccarthy , c. 1999 , , 512 , l63 weidenschilling , s. j. 1997 , icarus , 127 , 290 weintraub , d. a. , saumon , d. , kastner , j. h. , & forveille , t. 2000 , , 530 , 867 wenrich , m. l. , & christensen , p. r. 1996 , , 101 , 15921 werner , m. w. , et al .
2004 , , 154 , 1 white , r. j. , ghez , a. m. , reid , i. n. , & schultz , g. 1999 , , 520 , 811 wilner , d. j. , ho , p. t. p. , kastner , j. h. , & rodrguez , l. f. 2000 , , 534 , l101 wood , j. a. 1996 , meteoritics & planet .
, 31 , 641 zubko , v. g. , mennella , v. , colangeli , l. , & bussoletti , e. 1996 , , 282 , 1321 cccccccc coku tau/4 & m1.5@xmath73 & 3.0@xmath73 & 3720@xmath73 & 0.61@xmath73 & 0.5@xmath74 & 0.5@xmath75 & 0.001 + dm tau & m1@xmath76 & 0.5@xmath76 & 3720@xmath76 & 0.25 & 0.65@xmath76 & 50@xmath76 & 0.077 + gm aur & k5@xmath76 & 1.2@xmath76 & 4730@xmath76 & 0.83 & 1.2@xmath76 & 90@xmath76 & 0.075 + tw hya & k7@xmath77 & 0@xmath78 & 4009@xmath77 & 0.25@xmath77 & 0.7@xmath74 & 30@xmath79 & 0.043 + fm tau & m0 & 0.69 & 3850 & 0.32 & 0.6@xmath74 & 2@xmath75 & 0.003 + ip tau & m0 & 0.24 & 3850 & 0.43 & 0.6@xmath74 & 3@xmath75 & 0.005 + gg tau aa & k7@xmath80 & 0.8@xmath80 & 4060@xmath81 & 0.84@xmath80 & 0.78@xmath80 & & + gg tau ab & m0.5@xmath80 & 3.2@xmath80 & 3850@xmath81 & 0.71@xmath80 & 0.68@xmath80 & & + gg tau ba & m5@xmath80 & 0.55@xmath80 & 3050@xmath80 & 0.08@xmath80 & 0.12@xmath80 & & + gg tau bb & m7@xmath80 & 0@xmath80 & 2820@xmath80 & 0.02@xmath80 & 0.044@xmath80 & & + hen 3 - 600 a & m3e@xmath77 & 0.7@xmath82 & 3350@xmath77 & 0.2@xmath77 & 0.5@xmath74 & & + fn tau & m5 & 1.35 & 3240 & 0.5 & 0.28@xmath74 & 4@xmath83 & 0.014 + v410 anon 13 & m5.75@xmath84 & 5.8@xmath84 & 3000@xmath84 & 0.077@xmath84 & 0.1@xmath84 & & + cy tau & m1 & 0.1 & 3720 & 0.47 & 0.5@xmath74 & 6@xmath75 & 0.012 + | the 814@xmath0 emission spectra of 12 t tauri stars in the taurus / auriga dark clouds and in the tw hydrae association obtained with the infrared spectrograph ( irs ) on board _
spitzer _ are analyzed .
assuming the 10@xmath0 features originate from silicate grains in the optically thin surface layers of t tauri disks , the 814@xmath0 dust emissivity for each object is derived from its spitzer spectrum .
the emissivities are fit with the opacities of laboratory analogs of cosmic dust .
the fits include small nonspherical grains of amorphous silicates ( pyroxene and olivine ) , crystalline silicates ( forsterite and pyroxene ) , and quartz , together with large fluffy amorphous silicate grains .
a wide range in the fraction of crystalline silicate grains as well as large silicate grains among these stars are found .
the dust in the transitional - disk objects coku tau/4 , gm aur , and dm tau has the simplest form of silicates , with almost no hint of crystalline components and modest amounts of large grains .
this indicates that the dust grains in these objects have been modified little from their origin in the interstellar medium .
other stars show various amounts of crystalline silicates , similar to the wide dispersion of the degree of crystallinity reported for herbig ae / be stars of mass @xmath12.5@xmath2
. late spectral type , low - mass stars can have significant fractions of crystalline silicate grains .
higher quartz mass fractions often accompany low amorphous olivine - to - amorphous pyroxene ratios .
it is also found that lower contrast of the 10@xmath0 feature accompanies greater crystallinity . 0 _ |
phase - change materials ( pcms ) constitute a class of semiconductors characterized by two allotrope phases , a crystalline and an amorphous one , having distinct physical properties @xcite .
such materials , discovered in the late 60s @xcite , are already used in rewritable optical discs , such as dvds and blu - ray discs .
they are also very promising to realize fast , non - volatile electronic memories ( pcram ) .
because of the variety of pcms a rigorous chemical definition is absent , although design schemes for some ternary compounds are well established @xcite . a practical requirement for a pcm , as underlined in ref .
@xcite , is that the switching between the two phases must be reversible , efficient and repeatable , due to obvious technological reasons .
benchmark pcms are alloys composed of ge , sb and te along the so called gete - sb@xmath0te@xmath1 pseudobinary line thus denoted as `` gst '' @xcite .
although the potential for applications of pcms spans from dynamic memories @xcite to display fabrication @xcite , at present the limits for the use of gsts are the speed and the power required to switch between the two phases , being respectively of approximately 100 ns and 100 @xmath3w @xcite .
hence a huge effort is devoted to find new solutions , and nanoscale structures seem to offer significant advantages .
@xcite . following this trend ,
phase - change heterostructures had been produced , showing to function with improved performances @xcite . here
we focus on chalcogenide superlattice ( csl ) films made by high temperature deposition of alternating nm - size layers of gete and sb@xmath0te@xmath1 @xcite .
+ prior to investigate the properties of csls , we review the charge carrier transport @xcite and the structural @xcite properties of gst compounds .
the crystal phase of gsts is characterized by a strong dependence of the resistivity upon annealing treatment @xcite .
in fact , by increasing the temperature , the system undergoes an insulator - metal transition ( imt ) at 275@xmath4c , due to disorder reduction .
@xcite density functional theory calculations showed that the insulating state is caused by the localization of charge carriers in vacancy - rich areas , and the transition to the metallic state happens when vacancies reconfigure into ordered layers @xcite .
the presence of vacancy layers in the high temperature annealed phase of gsts was proven by high - resolution tem and electron diffraction @xcite .
+ the lattice structure of the crystalline ge - sb - te based csl is similar , by growth design , to the high temperature gst phase . as mentioned before csl
are formed by high temperature deposition although lower than the imt of gete and sb@xmath0te@xmath1 layers .
one might consequently expect that the as grown structure is characterized by well separated layers of pure gete and sb@xmath0te@xmath1 .
recent tem and exafs experiments @xcite have shown that the structure is indeed layered , but gete is not actually pure since it is intercalated into a sb@xmath0te@xmath1 quintuple layer .
please note that the vacancy layers , when two - dimensionally extended , are labelled van der waals gaps .
@xcite the similarity between the high - temperature - annealed gsts @xcite and the csls , besides the structure , is corroborated by the measured resistivity of the latter that decreases with increasing temperature , suggesting a metallic type conductivity @xcite . nonetheless , it is not obvious whether also the dielectric properties of csls are analogous to those of gst compounds @xcite .
for instance one of the peculiar characteristics of crystalline gst is the high value of the static dielectric constant , which is evidence of resonance bonding .
@xcite measuring the dielectric function of cls will provide useful information either about the structure , or about the bonding character of the compounds .
the dielectric function of csls can be estimated by transmittance experiments , probing simultaneously the character of the interband transition and the density of free electrons .
+ in this article we compare the optical transmission behaviour of films grown by dc magnetron sputtering ( ms ) , with those grown by molecular beam epitaxy ( mbe ) used as reference samples .
transmission experiments are performed on films with thickness below 100 nm , to guarantee high film quality , and at normal incidence , to simplify the interpretation of the results .
our analysis supports recent results showing that the gete layers are not isolated but intercalated within a sb@xmath0te@xmath1 layer @xcite .
we explain how intermixing modifies the conductivity of the film .
in addition we show that it is possible to control the electronic transport properties of the csl not only by an annealing treatment , as in ref .
@xcite , but also by means of a specific stack design .
in fact , as we will see , the csl conductivity depends on the layer sequence , giving a new design pathway for the control of the material properties .
csl samples are grown by ms and mbe on sb - passivated si(111 ) surfaces of intrinsic 500 @xmath3m - thick wafers above 210 @xmath4c .
ms samples are films of 15 repetitions of [ gete(@xmath5 nm)/sb@xmath0te@xmath1(3 nm ) ] , where @xmath6 . for the mbe samples we use the intermediate n=2 case .
ms annealed ( ann ) samples at 250 @xmath4c for 30 minutes are distinguished from the ms as deposited ( asd ) ones .
transmission experiments in the energy range 0.061.23 ev were collected at sissi beamline on the elettra storage ring @xcite by a @xmath7 spectrometer .
the fitting function is determined for multilayer thin films optics @xcite .
figure [ tall ] shows the absorbance @xmath8ln@xmath9 , with @xmath10 and @xmath11 respectively the sample and the substrate transmission , as a function of the incoming photon energy for all the samples .
the data present few physically relevant features , common to all samples and regardless of growth technique and layer stack . in particular , we observe i ) the onset of an absorption edge below @xmath20.15 ev , increasing by decreasing the photon energy ; ii ) a minimum between 0.15 and 0.3 ev , and iii ) a broad feature rising above @xmath20.4 ev , increasing towards higher energies . in analogy to ref .
@xcite we associate the low frequency absorption component to free drude electrons , and the higher energy broad band to the onset of the valence - to - conduction absorption .
the absorption modulation of the data , particularly visible on the ms 33 annealed sample , is an artefact due to multilayer interference effects .
we account for interference effects by calculating the transmission coefficient @xmath12 including reflection losses .
+ the film dielectric function ( df ) is modelled with a sum of a drude and a single tauc - lorentz ( tl ) oscillator term @xcite : @xmath13 where @xmath14 with @xmath15 the photon energy , @xmath16 the plasma frequency and @xmath17 the scattering time .
the real part of @xmath18 is a negative lorentzian function centred at zero while the imaginary part is positive and diverges at zero .
the tauc - lorentz dielectric function is obtained by kramers - kronig integration ( @xmath19 ) from the imaginary part : @xmath20}\\ \end{array } \right . , \ ] ] where @xmath21 is the tl amplitude , @xmath22 is the tauc or optical gap , @xmath23 is the band - width and @xmath24 is the central frequency , representing the photon energy where the transition probability is at the maximum . .
b ) dielectric function resulting from the fit .
both the real ( red ) and the imaginary ( blue ) components of the df are shown also subtracting the drude contribution ( dashed ) . ] for the ( intrinsic ) silicon substrate we use @xmath25 .
we verified the assumptions made for the dielectric properties of the substrate by transmittance measurements on virgin wafers . regarding the calculation of @xmath26
we include multiple coherent propagation only for the thin csl film .
also the csl film thickness is a free parameter of the fitting procedure .
a typical fit result is shown in figure [ fitres]a .
the related df is plotted in figure [ fitres]b .
the misfit between the experimental and calculated data is due to si phonons ( the peak at @xmath20.3 ev ) and probably to the small contribution of interference effect from the substrate that have not been accounted for . in all cases
the fit function @xmath26 describes all the main features of the transmission data , thus the deduced parameters are meaningful .
+ the fit parameters for the tl oscillator and the drude term , for each investigated csl , are given in table [ tab_par ] .
+ obtained from the drude term of the fit .
a decrease of @xmath27 is observed with increasing gete content . ]
.paremeters resulting from the fit of eq .
( [ eps ] ) for all the measured samples .
the error on the fit procedure is shown in parenthesis , representing the deviation of the last digit .
the low - energy limit of the tl df , @xmath28 , is also computed . [ cols=">,^,^,^,^,^,^ " , ]
we first consider the tl component of the df , whose @xmath22(ev ) , @xmath24(ev ) , @xmath28 parameters are given in table [ tab_par ] .
the tauc gap lays below 0.5 ev .
@xmath24 , representing the maximum of the semiconductor optical absorption , has an average value of approximately 1.5 ev .
this value matches quite well with the experimentally obtained value of crystalline gete @xcite and gst @xcite .
moreover , the computed low frequency value of the real part of the tl df , @xmath28 , has an average value above 20 .
this @xmath28 high value is typical of the crystal phase of pcms @xcite , suggesting the presence of resonance bondings . in summary
the tl component of our csl is approximately independent on the annealing treatment .
the fact that annealing does not affect the as - deposited ( asd ) structures might suggest that the films undergo an immediate reorganization during growth , which is more effective than annealing itself .
@xcite the analysis of the free electrons in our csl supports this hypothesis .
+ te@xmath1 and gst materials .
the tauc gap @xmath22 in both gete and sb@xmath0te@xmath1 is lower with respect to gst compounds .
thermal promotion of valence band ( vb ) electrons to the conduction band ( cb ) is accordingly reduced in gsts , hence the free electron density and the conductivity are lower .
b ) formation of gst layers at the interface between gete and sb@xmath0te@xmath1 . ] the plasma frequency and the scattering time , resulting from the fitting procedure , are presented in table [ tab_par ] . while the scattering time ranges between 5 - 15 fs
just half with respect to copper at 300 k @xcite the plasma frequency consistently decreases with the increase of the stacking sequence 13 - 23 - 33 , both for the asd and ann samples .
the csl electron density @xmath29 , derived from @xmath16 , is shown in figure [ fin_par ] .
a decrease of @xmath27 is observed with the increase of nominal gete content in the film .
also the conductivity @xmath30 shows the same trend .
this behaviour can be explained either by an increased defectivity , by a lower conductivity of gete with respect to sb@xmath0te@xmath1 , or by an increase of the optical gap . yet , since the comparison acts within the ms grown samples , we expect the same growth quality and consequently the same concentration of defects .
in addition , pure gete has a _ higher _ free carrier concentration with respect to sb@xmath0te@xmath1 .
@xcite thus , the expected trend with stack sequence 13 - 23 - 33 should be opposite than what observed .
it follows that variations of @xmath27 must result from changes of the electronic band structure .
+ the intermixing of gete and sb@xmath0te@xmath1 layers rises the material optical gap , causing a reduction of the free carrier concentration .
figure [ scheme]a sketches the band diagrams for gete , sb@xmath0te@xmath1 and gsts , in analogy to refs @xcite .
the diagrams emphasize the differences between the optical gaps and the presence of degenerate electrons in the conduction band .
both gete and sb@xmath0te@xmath1 have a low tauc gap @xmath310.1 ev @xcite , while for gst @xmath32 0.4 ev @xcite . in first approximation
the tauc gap of the csl depends on the constituents volume fraction and on their respective gaps .
if gete and sb@xmath0te@xmath1 are ideally separated inside the csl the film is expected to have small gap and high carrier concentration , independent of the relative content of the two elements .
if instead gst is formed by the intermixing of the two , the material tauc gap must rise .
the more gst is formed , the higher the tauc gap ( approaching eventually the value of gsts ) and the lower the number of degenerate electrons .
this is consistent with the carrier density of the sequence 13 - 23 - 33 , shown in figure [ fin_par ] .
) for increasing number of gete units , and compared with the measured conductivity of the csls .
b ) conductivity of a film with stack sequence ( gete)@xmath33/(sb@xmath0te@xmath1)@xmath34 . the red dashed diagonal indicates the lowest conductivity region .
the black dashed horizontal lines represent the cut used to generate the graph a ) . ]
we extrapolate the trend observed in the present csl to a stack of general thickness , having ( gete)@xmath33/(sb@xmath0te@xmath1)@xmath34 as repeat unit , where @xmath35 .
this formula is applicable to all csl structures developed so far in literature : i ) superlattice - like ( ssl ) pcms , with @xmath36 @xcite ; ii ) interfacial phase change materials ( ipcm ) , with @xmath37 @xcite ; iii ) the case @xmath38 , studied in ref .
@xcite and iv ) our csls .
we assume our stack sequence corresponds to @xmath39 and @xmath40 , since 1 nm of gete along the ( 111 ) direction is composed by at least two bilayers @xcite .
the conductivity of a csl in the general case is calculated assuming that gete and sb@xmath0te@xmath1 layers intermix at the interface , forming gst as shown in figure [ scheme]b . in particular , we consider that the interface layer consists of a maximum of 2 sb@xmath0te@xmath1 and 4 gete layers , resulting in 1 thick layer of gst . then , given the bulk conductivities @xmath41 , @xmath42 , @xmath43 @xcite , the film conductivity is calculated as : @xmath44 where @xmath45 is the volume fraction of the @xmath46-th component ( gete , sb@xmath0te@xmath1 and gst ) . + in figure [ nm]a , the conductivity measured for all our samples
is compared with the conductivity @xmath47 , as a function of gete layer units @xmath5 . in agreement with experimental observations
the conductivity decreases from @xmath48 down to @xmath49 .
the strikingly different behaviour of the `` non - intermixing '' case , calculated simply via eq .
( [ sigma ] ) with @xmath50 , demonstrates once more that intermixing effects must be accounted for when working with superlattice structures . + for completeness the dependence of @xmath51 is shown in figure [ nm]b .
interesting to note the conductivity has a minimum along the diagonal @xmath52 up to @xmath53 that corresponds to the maximum possible volumetric formation of gst , hence forming the material with the lowest conductivity . for higher values of @xmath5 and @xmath54 the conductivity approaches the `` non - intermixing '' case , as also visible for high values of gete units in figure [ nm]a .
in this work we deduce important structural details of crystalline gete / sb@xmath0te@xmath1 superlattice deposited at high temperature ( @xmath55210 @xmath4c ) by studying the film interband transmission .
we observe that , at the interface between the gete and sb@xmath0te@xmath1 deposition layers , a crystalline gst compound is formed already during growth .
gst , having higher band gap with respect to its constituents , lowers the number of degenerate conduction electrons . by varying the respective number of building block layers it is possible to control the percent of gst in the film , and consequently the film conductivity , with no need of annealing treatment . in the future
it would be interesting to investigate the dynamics deriving from the intermixing and ion diffusion by depositing two single thick layers of gete and sb@xmath0te@xmath1 and studying the interface effects at different deposition and annealing conditions .
this work was supported by eu within fp7 project pastry [ ga 317764 ] .
ac would like to acknowledge roberta ciprian for useful comments .
48ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1002/adma.201004255 [ * * , ( ) ] link:\doibase 10.1002/adfm.201500826 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.21.1450 [ * * , ( ) ] @noop _
_ , oxford : university press ( ) http://dx.doi.org/10.1038/nmat2330 [ * * , ( ) ] @noop * * ( ) @noop * * ( ) http://dx.doi.org/10.1038/nmat2009 [ * * , ( ) ] http://dx.doi.org/10.1038/nmat1359 [ * * , ( ) ] link:\doibase 10.1038/nature13487 [ * * , ( ) ] link:\doibase 10.1145/1555815.1555758 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1039/c5nr04530d [ * * , ( ) ] \doibase doi:10.1002/admi.201300027 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.22.1065 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.76.235201 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1002/adfm.201500848 [ * * , ( ) ] @noop * * ( ) http://dx.doi.org/10.1038/srep22353 [ * * , ( ) ] link:\doibase 10.1007/978 - 3 - 642 - 20018 - 2_31 [ _ _ ] , edited by ( , , ) chap .
, pp . @noop link:\doibase 10.1103/physrevb.8.660 [ * * , ( ) ] http://stacks.iop.org/0022-3719/12/i=21/a=013 [ * * , ( ) ] link:\doibase 10.1143/jjap.28.1407 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.081204 [ * * , ( ) ] \doibase http://dx.doi.org/10.1038/nmat2226 [ * * , ( ) ] link:\doibase 10.1364/josab.24.000959 [ * * , ( ) ] @noop @noop * * ( ) \doibase http://dx.doi.org/10.1063/1.4932666 [ * * , ( ) , http://dx.doi.org/10.1063/1.4932666 ] link:\doibase 10.1021/acs.cgd.5b01714 [ * * , ( ) ] @noop _ _ link:\doibase 10.1103/physrevb.52.16321 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.4928630 [ * * , ( ) , http://dx.doi.org/10.1063/1.4928630 ] link:\doibase 10.1103/physrev.172.779 [ * * , ( ) ] @noop * * ( ) link:\doibase 10.1007/bf00324584 [ * * , ( ) ] http://stacks.iop.org/1347-4065/44/i=10r/a=7340 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.2959818 [ * * , ( ) , http://dx.doi.org/10.1063/1.2959818 ] \doibase http://dx.doi.org/10.1063/1.2181191 [ * * , ( ) ] | the extraordinary electronic and optical properties of the crystal - to - amorphous transition in phase - change materials led to important developments in memory applications .
a promising outlook is offered by nanoscaling such phase - change structures . following this research line ,
we study the interband optical transmission spectra of nanoscaled gete / sb@xmath0te@xmath1 chalcogenide superlattice films .
we determine , for films with varying stacking sequence and growth methods , the density and scattering time of the free electrons , and the characteristics of the valence - to - conduction transition .
it is found that the free electron density decreases with increasing gete content , for sub - layer thickness below @xmath23 nm .
a simple band model analysis suggests that gete and sb@xmath0te@xmath1 layers mix , forming a standard gesbte alloy buffer layer .
we show that it is possible to control the electronic transport properties of the films by properly choosing the deposition layer thickness and we derive a model for arbitrary film stacks . |
during messenger s third flyby of mercury , a 290-km - diameter peak - ring ( double - ring ) impact basin , centered at 27.6@xmath0 n , 57.6@xmath0 e , was discovered and subsequently named rachmaninoff . in terms of size and morphology ,
the rachmaninoff basin closely resembles the 265-km - diameter raditladi peak - ring basin , located at 27@xmath0 n , 119@xmath0 e west of the caloris basin , that was discovered during messenger s first flyby @xcite .
the image - mosaic of rachmaninoff and its ejecta has a spatial resolution of 500 m / pixel and it is derived from images obtained by messenger s mercury dual imaging system ( mdis ) narrow - angle camera @xcite , while raditladi basin and surrounding areas were imaged at 280 m / pixel .
both basins and surrounding areas were also imaged with a set of 11 filters of the mdis wide - angle camera ( wac ) , whose wavelengths range from 430 to 1020 nm @xcite .
these images were used to obtain color maps with a resolution of about 5 km / pixel and 2.4 km / pixel for rachmaninoff and raditladi , respectively .
+ the two basins appeared to be remarkably young because of the small number of impact craters seen within their rims @xcite .
for this reason it has been argued that they were likely formed well after the end of the late heavy bombardment of the inner solar system at about 3.8 ga @xcite . in particular , for raditladi it has been pointed out that the basin could be as young as 1 ga or less @xcite .
+ interestingly , both basin floors are partially covered by smooth plains . in the case of rachmaninoff ,
an inner floor filled with spectrally distinct smooth plains has been observed and this , combined with the small number of overimposed craters , implies a volcanic origin @xcite .
the estimate of the temporal extent of the volcanic activity and , in particular , the timing of the most recent activity may represent a key element in our understanding of the global thermal evolution of mercury , and helps to constrain the duration of the geologic activity on the planet in light of the new data provided by messenger .
moreover , raditladi may be the youngest impact basin discovered on mercury so far , and therefore it is important to understanding the recent impact history of the planet .
+ for all these reasons , the age determination of rachmaninoff and raditladi basins and their geologically different terrains is of a great interest . in this paper , we will present a revised mercury crater chronology , and show how to take into account for the crustal properties of the target ( @xmath1 3 ) .
this chronology will be then applied to rachmaninoff and raditladi basins ( @xmath1 4 ) .
in this paper we date the raditladi and rachmaninoff basin units by means of the the model production function ( mpf ) chronology of mercury @xcite .
this chronology relies on the knowledge of the impactor flux on mercury and on the computed ratio of impactors between mercury and the moon .
the absolute age calibration is provided by the apollo sample radiometric ages .
the crater scaling law enables computation of the crater size - frequency distribution ( sfd ) using a combination of the impactor sfd and the inferred physical properties of the target .
the computed crater sfd per unit surface and unit time is the so - called mpf .
the present model involves several improvements with respect to the model presented in @xcite and @xcite , thus it will be described in detail in the next sections .
+ in the following analysis , we use the present near - earth object ( neo ) population as the prime source of impactors .
this assumption is justified by the presumably young ages ( i.e. low crater density ) of the terrains studied in this paper . in particular
, we use the neo sfd as modeled by @xcite .
this neo sfd is in good agreement with the observed neo population , fireballs and bolide events ( see * ? ? ?
* for further details ) .
+ concerning the crater scaling law , we adopted the so - called pi - scaling law in the formulation by @xcite .
unlike previous approaches , our methodology explicitly takes into account the crustal properties of the target .
in fact , surfaces react differently to impact processes , depending on the bulk density , strength and bulk structure of the target material .
these latter parameters are taken into account by the scaling law , and are tabulated for several materials like cohesive soils , hard - rock and porous materials ( e.g. * ? ? ?
* ; * ? ? ?
* ) . on a planetary body
, terrain properties may vary from place to place according to the local geological history and as a function of the depth in the target crust .
therefore , impacts of different sizes taking place on a particular terrain may require different estimates of the target properties .
+ the pi - scaling law allows computation of the transient crater diameter ( @xmath2 ) as a function of impact conditions and target properties , and reads : @xmath3^{-\frac{\mu}{2+\mu } } \label{hh}\ ] ] where @xmath4 is the target gravitational acceleration , @xmath5 is the perpendicular component of the impactor velocity , @xmath6 is the projectile density , @xmath7 and @xmath8 are the density and tensile strength of the target , @xmath9 and @xmath10 depend on the cohesion of the target material and @xmath11 on its porosity and @xmath12 for cohesive soils , while @xmath13 and @xmath14 for rocks .
@xmath11 has been set to 0.4 in all cases @xcite . ] .
therefore , the nature of the terrain affects the crater efficiency and the functional dependence of the crater size with respect to the input parameters ( e.g. impactor size and velocity ) .
equation [ hh ] accounts both for the strength and gravity regimes , allowing a smooth transition between the two regimes .
the impactor size ( @xmath15 ) for which we have the transition between the two regimes is determined by equating the two additive terms in equ .
[ hh ] , therefore : @xmath16 the transient crater diameter is converted into final crater diameter ( @xmath17 ) according to the following expressions : @xmath18 where @xmath19 is the observed simple - to - complex transition crater diameter , which for mercury is 11 km @xcite . the conversion between transient crater to final crater is rather uncertain and several estimates are available ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
here we have used the factor 1.3 from transient to final simple craters . for the complex craters , we use the expression proposed by @xcite , where the constant factor has been set to 1.40 in order to have continuity with the simple crater regime ( @xmath20 ) .
we note that the effects of the material parameters on the transient crater size depend on whether a crater is formed in the strength or gravity regime . for the strength regime , @xmath21 or @xmath22 , while @xmath23 or @xmath24 according to the values for @xmath10 and @xmath11 given above . in the gravity regime , @xmath25 . in all cases ,
the dependence of @xmath17 on both @xmath7 and @xmath8 is mitigated by the low exponents .
the application of the crater scaling law is not straightforward , since the physical parameters of the terrains are poorly constrained for mercury and the crater scaling law has been derived for idealized uniform target properties .
so far , no detailed and systematic study has been performed to develop a crater scaling law for a layered target ( e.g. * ? ? ?
* ; * ? ? ?
* ) , although numerical modeling of terrestrial craters has shown that the target layering plays an important role in the cratering process ( e.g. * ? ? ?
* and references therein ) .
on the other hand , we think that it is worth attempting to simulate a more realistic situation instead of using the same average values for craters whose sizes can vary by order of magnitudes and consequently involved different layers of a planetary crust . in this context
, geological analysis of the terrains can provide valuable information , at least to constrain the surface properties of the target .
+ in this work , we assumed that the density and strength of mercury varies as a function of the depth , in analogy to that inferred for the moon ( * ? ? ? * and references therein ) . in figure [ den_str ] ( left panels ) the assumed density and strength profiles
are indicated .
these profiles are consistent with the upper lunar structure @xcite , and were adopted also for mercury . in particular , we have considered a more or less fractured upper crust on top of a bulk silicic lower crust which in turn overlays a peridotitic mantle .
however , it must be emphasized that the depths at which these layers occur may vary from place to place ( see @xmath1 3 ) .
+ for each impactor size , we have assigned average values for the target density and strength . over a wide range of parameters ,
the transient crater radius ( @xmath26 ) is about @xmath27 times larger that the impactor radius @xmath28 and the depth of the crater is typically between one - fourth and one - third of the crater size @xcite .
thus the thickness of the excavated material is roughly between @xmath29 impactor radii .
here we have adopted an intermediate value , namely averaging the density and strength up to a depth of @xmath30 ( see fig .
[ den_str ] , right panels ) given the limited variation in the density and strength profile , the choice of the actual depth to average the density and strength for a given impactor radius has a low influence ( @xmath31 ) on the scaling law . +
in addition to the density and strength profiles , we also consider a transition of the crater scaling law ( from cohesive - soil to hard - rock ) according to the size of the impactor .
in fact , the density and strength profiles shown in fig .
[ den_str ] describe a material of increasing coherence for increasing depth .
this is the result of the continuous bombardment of planetary surfaces that produces comminution and fracturing of the upper crustal layers gradually deacreasing with depth as observed in seismic profiles of the lunar crust underneath the mare cognitum @xcite . in this respect ,
craters that affect only the upper fractured layers form in the cohesive - soil regime , while larger ones in hard - rock regime .
therefore the depth of the transition ( @xmath32 ) from the superficial fractured layer to the unfractured lower crust is an important parameter .
the depth ( and therefore the crater size ) at which the transition from one regime to the other occurs can vary from place to place .
for instance , the thickness of the cohesive layer may be only of a few meters on recent lunar mare material @xcite , while it is expected to be of several kilometers on the highlands . in the examples of fig .
[ den_str ] , it is assumed that @xmath33 km @xcite . + the details of the transition are not easy to model .
a simplified study of impact processes on a layered target was performed by @xcite .
they simulated a two - layer structure , formed by a loose , granular layer on top a more competent material .
it was observed that the craters had the usual shapes for diameters less than about 4 times of the top layer thickness .
larger craters developed central mounds , flat floors and concentric rims indicating the presence of the underlying layer . according to these results , we simplify the problem by considering a sharp transition in the crater scaling law , at @xmath34 .
this implies a transition as a function of impactor radius at @xmath35 .
the effect of the transition in the scaling law is reported in fig .
[ sl ] for two depths of transitions .
note that the position of the sharp transition varies according to the depth assumed . in this paper
, we use an intermediate value and set the transition at @xmath36 . in a more realistic situation , a gradual transition
should be predicted given that the target gradually changes its properties as function of the depth .
therefore , our simplified model is not expected to be accurate close to the transition region , nevertheless we believe it provides a reasonable way to approach the cratering scaling law for a layered target . + the neo population and the crater scaling described in the previous section are used to derive the mpf per unit time ( see fig .
[ mpf ] ) .
the main outcome of our model is that the adopted transition in the crater scaling law results into a `` s - shaped '' feature ( or flexure ) in the mpf .
the position of such feature , which is determined by @xmath32 , is not known a - priori .
however , as discussed in @xcite , in some cases @xmath32 can be constrained by the shape of the observed crater sfd .
furthermore , the geological analysis of the terrains can help to derive the expected range of variation for @xmath32 .
for instance , lava emplacements may partly strengthen or even completely replace the pre - existing fractured layer .
hence in this latter cases the fractured horizon can be confined within a very thin regolith cover , negligible for our calculation ( @xmath37 ) . for young units with poor crater statistics
, the choice of @xmath32 may affect the age determinations by up to a factor of 3 - 4 .
thus , in order to derive a more accurate age estimate , it is of paramount importance to adapt the crater production function to the nature of the terrains investigated .
+ the absolute age is given by the lunar chronology , which expresses the lunar crater cumulative number at 1 km ( @xmath38 ) as a function of time ( @xmath39 ) , using the following equation : @xmath40 where @xmath41 , @xmath42 , @xmath43 @xcite .
the mpf function at a time @xmath39 is given by : @xmath44 the mpf(@xmath39 ) is used to derive the model cratering age by a best fit procedure that minimizes the reduced chi squared value , @xmath45 .
data points are weighted according to their measurement errors .
the formal errors on the best age correspond to a 50% increase of the @xmath45 around the minimum value .
+ it must be realized that the formal statistical error on the model age only reflects the quality of the crater sfds . on top of that
, other sources of uncertainties are present .
they stem from the uncertainties involved in the physical parameters used in the model , although we want to stress here that the model ages are not very sensitive to details of the density and strength profile ( see @xmath1 3.1 of * ? ? ?
a more important issue is the applicability of the present neo population in the past .
the @xmath46 chronology function assumes a linear dependence with time in the last @xmath47 ga , corresponding to an impactor population in a steady state . on the other hand
, dynamical studies of recent main belt asteroid family formation suggested that the present neo flux may be higher that the average steady flux by a factor of 2 @xcite .
this result also agrees to what found by cratering studies of young lunar terrains ( @xmath48 ga ; e.g. * ? ? ?
* ) . concerning the layering structure , as described above
, the choice of @xmath32 might affect the age estimate by a factor of 3 - 4 at most .
the uncertainty due to the layering is , nevertheless , typically present only for young terrains where the crater sfd has a limited range of crater dimensions .
the layering affects the specific shape of the crater sfd , and has to be evaluated case by case , as it will be discussed in detail in the following sections .
finally , it must be noted that wavy features in the crater sfd can be due also to other processes than layering , like for instance partial crater obliteration due to subsequent lava flows .
therefore , the nature of the s - shaped feature must be constrained as much as possible by geological analysis in order to achieve a more reliable age determination .
geological maps of the rachmaninoff and raditladi basins were constructed considering both floors and ejecta . for the floor terrains ,
the geological units were identified on the base of their different surface morphologies and spectral characteristics ( i.e. , albedo ) , along with an analysis of their stratigraphic relationships .
the ejecta units , surrounding the basins , were outlined considering exclusively the area of continuous ejecta blankets , which are easily detectable thanks to their characteristic hummocky surface .
the geological maps also take into account tectonic features affecting the areas .
+ crater age determination is based on the primary craters , i.e. those formed by impacts with objects in heliocentric orbits .
hence , a crucial point in assessing age by crater counts is to identify and avoid secondary craters .
most of the secondaries are recognizable because they are directly related to their primaries ( e.g. , the secondaries are arranged in radial patterns around the primary ) , or occurr in loops , cluster and chains .
the contribution of far - field secondaries , which are normally not distinguishable from primary craters ( e.g. * ? ? ? * ) , has been neglected .
although this is a reasonable assumption for rachmaninoff and raditladi basins given their low crater densities and thus their presumable relatively young ages , mpf model ages may overestimate real ages .
+ the rachmaninoff basin is surrounded by a continuous ejecta blanket and includes an interior peak ring structure , about 136 km in diameter , with extended smooth plains filling its floor .
most of the basin walls are modified into terraces .
several different geological units have been distinguished inside the floor on the basis of their different relative albedo and surface texture ( fig .
[ rach_geo ] ) .
the inner smooth plains are mostly within the peak ring except in the southern quadrant of rachmaninoff , where the smooth plains cover or embay the peak ring structure and some of the annular units within the rim .
this observation suggests an origin of volcanic emplacement . in the wac enhanced - color images these plains show a yellow to reddish tone which stands out from the darker and bluer color of the other units within the basin and surrounding regions
this clearly supports a different composition and origin of these inner smooth plains .
several discontinuous and concentric troughs possibly due to the uplift and extension of the basin floor , affect the area enclosed by the peak ring @xcite and have been interpreted to be graben .
the annular region between the peak ring and the rim basin includes seven different units .
the most prominent is made up of bright materials , apparently younger than all the other units and possibly related to explosive volcanism @xcite .
peak ring and terrace material boundaries stand out for their relief , whereas hummocky , dark , irregular and annular smooth plains do not show unequivocal stratigraphic relationships with each other @xcite .
this suggests an almost coeval origin of these units that may consist of impact melts and breccias .
this is furthermore confirmed by the wac images , where annular units do not reveal any color variations and are characterized by uniform blue color similar to the surrounding terrain . to shed more light on the origin of the floor material we dated the rachmaninoff basin using crater statistics of annular units and inner plains separately .
bright material was neglected in the crater counts due to its limited extent .
craters in the ejecta blanket were counted as well ( fig . [ rach_geo ] ) .
+ easily recognizable secondary craters ( either elliptical in shape or arranged in loops and chains ) have not been detected within the rachmaninoff inner plains but numerous pits up to 3.5 km in diameter have been found in close proximity to the concentric grabens .
these features are very unlikely to be impact craters and in our interpretation are most probably of tectonic ( structural pits , fault bounded depressions , en - echelon structures ) and/or volcanic ( more or less irregular vents ) origin ( fig .
[ rach_ex ] , panels a and b ) . for this reason
their counts were neglected for the purposes of age determination .
clusters and chains of secondaries with irregular and elliptical shapes were recognized in the western sector of the annular units and appear to be directly related to a nearby 60 km primary peak crater overlying the rachmaninoff ejecta ( fig .
[ rach_ex ] , panels c and d ) .
self secondaries are numerous within the ejecta blanket .
+ the resulting crater count statistics are reported in table [ tab ] where for each terrain , all crater - like features , bonafide craters , secondary craters and endogenic ( namely volcanic or tectonic ) features are listed .
it is interesting to compare the sfds of all the counts ( fig .
[ rach_sfd ] ) . according to our best interpretation of the detected features ,
the inner plains contain more endogenic features than bonafide craters .
hence , in our opinion the uneven distribution in r plots of crater - like features smaller than 4 km , which is generally attributed to the effect of far field secondaries on mercury @xcite , are in this case dominated by tectonic and volcanic features . for the annular units , both the secondary crater and endogenic sfds have steeper slopes with respect to the bonafide crater sfd
, moreover they are limited to features smaller than 4 - 5 km .
hence , for the annular units as well as the inner plains , the identification of endogenic features is clearly very important since it heavily affects the final bonafide crater sfd . on the ejecta ,
most of the crater - like features appear to be self secondaries mostly arranged into clusters and chains and/or with an elliptical shape .
all the terrains were dated using their bonafide crater sfds .
+ . statistics of all the features detected on rachmaninoff and raditladi basin floors and ejecta .
`` all '' indicates all crater - like features , `` bon '' the bonafide craters , `` sec '' secondary craters ( which includes chain , cluster and elliptical craters ) , `` end '' endogenic ( volcanic and tectonic ) features .
pl . '' and `` ann . un
. '' stand for inner plain and annular unit , respectively .
[ cols="<,^,^,^,^,^,^ " , ] the mpf fits of the observed crater sfds are shown in figure [ rach_mpf ] .
the lower panel shows the distribution of bonafide primary impact craters detected on the ejecta blanket .
a remarkable feature in the crater sfd of the ejecta is the presence of a flexure point at about @xmath49 km .
the actual shape of the bonafide crater sfd is partially due to the feature selection .
nevertheless , we think that at the large crater sizes relevant here , our selection is reliable and , consequently , the observed flexure point is likely a real feature possibly reflecting a layered target with an upper weak horizon .
hence the mpf best fit is achieved with @xmath50 km and gives a model age of @xmath51 ga .
[ rach_mpf ] ( upper panel ) the bonafide crater sfd on inner plains and annular units are shown .
note that , unlike the ejecta , both cases do not show the presence of a flexure point .
this may be due to a real absence of an upper weak horizon or to the lack of large craters that would have otherwise allowed to retrieve information on the geomechanical properties of the deep crustal layers .
the annular units are composed of breccias more or less welded by impact melts which can have only partially strengthened the fractured material either pre - dating or originating from the rachmaninoff impact .
hence , it is reasonable to assume at least the same @xmath32 of the crust beneath the ejecta .
this leads to a model age of @xmath52 ga , which is consistent with the model age of the ejecta and likely dates the rachmaninoff impact event .
+ the inner plains are characterized by much poorer statistics within a small range of diameters , therefore the crater sfd can not be used to infer @xmath32 .
nevertheless , geological analysis suggest that the inner plains are younger volcanic flows on the basis of their different albedo , color and overlapping relationship with respect to the unit emplaced between the peak - ring and the basin rim @xcite .
this would make possible also the scenario in which the former fractured horizon , either pre - dating or originating from the impact itself , was completely hardened by the rising magmas and emplacement of lava fields ( fig .
[ rac_scenario]a ) . in this case
, the mpf acceptably fits the bonafide crater size distribution giving a model age of @xmath53 ga ( fig . [ rach_mpf ] ) .
by contrast the fit would be very poor if the all crater - like features would be taken into account .
this is not surprising considering the strong contribution we infer that tectonic and volcanic features have on the inner plains statistics .
another possible scenario is that the magmatic activity within rachmaninoff was unable to totally strengthen the upper weak layer comprising of fractured material originated by the impact itself or inherited by primordial events .
this could be due to either weakly sustained volcanism , which emplaced a thin volcanic sequence on top of a fractured material , and/or , a magma influx concentrated along few well defined conduits within a still fractured crust underneath the basin ( fig .
[ rac_scenario]b ) . in this case
we also computed the model age using @xmath50 km as for the annular units and ejecta , obtaining a value of @xmath54 ga . in both cases ( upper weak crustal layer absent or preserved ) the inner plains turn out to be remarkably young , and
demonstrate that a recent volcanic activity occurred within the basin .
+ raditladi contains an interior peak - ring structure 125 km in diameter and its walls appear to be degraded , with terraces most pronounced within the north and west sides of the rim @xcite .
a continuous ejecta blanket with no visible system of rays surrounds the basin and extends up to 225 km from the basin rim ( fig .
[ rad_geo ] ) .
the floor is partially filled with smooth , bright reddish plains material that clearly embays the rim and the central peak ring @xcite .
the northern and southern sectors of the basin floor consist of dark , relatively blue hummocky plains material confined between the rim and the peak ring .
troughs are found close to the center of the basin arranged in a partially concentric pattern , @xmath5570 km in diameter , and are interpreted either as graben resulting from post - impact uplift of the basin floor , or as circular dikes possibly representing fissural feeding vents @xcite .
floor material was subdivided into two different units following @xcite : smooth and hummocky plains .
smooth plains may have a volcanic origin , as appears to be the case for plains in the nearby caloris basin @xcite , however no clear stratigraphic relation with the hummocky plains has been found , suggesting that all the different terrains within raditladi basin may be coeval and directly related to the impact @xcite . with respect to the ejecta area ,
we have selected the hummocky continuous ejecta blanket surrounding the basin .
+ we performed a crater count of the inner plains within the peak ring and the annular units enclosed between the basin rim and the peak ring .
counts were also performed on the ejecta blankets .
+ within the inner plains , numerous small graben - related pits up to 5 km in diameter are identified ; as for the case of rachmaninoff , they are most probably tectonically - originated features and/or volcanic vents ( fig .
[ rad_ex ] , panels a and b ) .
specifically , two peculiar pits in the northern inner plains were interpreted as volcanic vents for the dark material on the crater floor @xcite .
secondary craters have been found associated to a 23-km crater within the inner plains ( fig .
[ rad_ex ] , panels a and c ) .
some secondaries are also present on the annular plains , whereas the ejecta blanket is characterized by numerous self secondary craters , occurring mainly in clusters and chains .
figure [ rad_geo ] shows the bonafide craters .
the statistics of all the identified features are reported in table [ tab ] , whereas the corresponding sfds are shown in fig .
[ rad_sfd ] .
+ the cumulative bonafide crater sfds for different terrains of the raditladi basin are shown in fig .
[ rad_mpf ] , along with the mpf model ages .
+ the measured crater sfd on the ejecta blanket shows a flexure at about @xmath56 km .
the position of the flexure is well above the size of craters that can no longer be distinguished because of the image resolution , nevertheless the contribution of secondary craters at these crater sizes might be important . in the assumption that this feature is due to the layering , the best fit achieved for @xmath57 km gives a model age of @xmath58 ga .
the best fit for the annular units is achieved using @xmath59 km and the resulting model age is @xmath60 ga .
these values are consistent with both the layering and model age inferred for the ejecta blankets .
for this reason the basin formation can be reliably fixed at around 1.1 - 1.3 ga in accordance with the age suggested by @xcite on the basis of a relative - chronology approach . as for the rachmaninoff basin ,
the poor statistics and the limited range of diameters imply that the inner plains sfd can not be used to constrain @xmath32 .
we derived the model age with both the same @xmath32 of the annular units ( @xmath59 km ) , and a negligible thickness ( @xmath37 ) , obtaining @xmath61 ga and @xmath60 ga , respectively . the former model age leads to a paradox given that the annular units are certainly coeval with the basin formation and , consequently , must only be older or of the same age as the smooth plains .
hence , the most reliable result for the inner plains is to consider a solid material yielding a crater retention age of about 1.1 ga . the solid material could be due to an emplacement of lavas soon afterward the impact leading to a complete hardening of the fractured and brecciated material within the basin ( fig .
[ rad_scenario]a ) .
this interpretation is consistent with the presence of volcano - tectonic features within the basin but may conflict with both the absence of distinctive color variations of the inner plains with respect to the surrounding areas and their unclear stratigraphic relationship with the annular units .
alternatively , a great amount of impact melts able to completely harden the impact breccias may explain the derived crater retention age ( fig .
[ rad_scenario]b ) .
mpf crater chronology has been applied to date the rachmaninoff and raditaldi basins .
age assesment has been performed taking into account target rheological layering and using the present neo population as the prime source of impactors .
+ our results demonstrate that the volcanic activity within the rachmaninoff basin interior significantly post - dates the formation of the basin .
the basin itself probably formed about 3.6 ga ago , whereas the volcanic inner plains may have formed less than 1 ga ago .
therefore , mercury had a prolonged volcanic activity , which possibly persisted even longer than on the moon , where the youngest detected nearside flows ( on oceanus procellarum ; * ? ? ?
* ) are about 1.1 ga old . on the other hand
, the raditladi basin has an estimated model age of about 1 ga and no firm indication that the inner plains formed more recently than the basin itself .
hence , these plains may be due to either huge volumes of impact melts or lavas emplaced soon afterward the basin formation . in the latter case , which is not clearly supported by the stratigraphic observations
, volcanism might have been triggered by the impact itself .
+ this work also shows the role of target properties in deriving the age of a surface .
where such properties are neglected , as in traditional chronologies ( e.g. * ? ? ?
* ) , the crater production function may be unable to accurately reproduce the observed crater sfd and/or to provide a consistent age for nearby terrains .
the following examples serve to illustrate this point : the rachmaninoff ejecta bonafide crater sfd shows an s - shaped feature that , according to our best knowledge , can not be ascribed to processes other than a layered target ; the raditladi inner plains have an higher density of craters than the annular units , implying a paradoxically older age for the interior plains if the inner and outer plains had the same material properties .
+ the derived ranges of ages for raditladi basin imply that its formation occurred long after the late heavy bombardment ( @xmath62 ga ) , at a time when the primary source of impactors was a neo - like population .
this conclusion also likely applies to rachmaninoff basin , even if it can not be excluded that it was formed during or prior the late heavy bombardment .
the neos average impact velocity on mercury is about 42 km / s @xcite . considering a most probable impact angle of @xmath63 ,
the projectiles responsible for rachmaninoff and raditladi formation should have had diameters in the range 14 - 16 km ( see fig .
[ sl ] ) . in the present neo population , bodies
are quickly replenished -in time scales of tens of myrs- mainly from the main belt via slow orbital migration into major resonances .
such a migration , due mainly to yarkowsky effect , is size dependent and is negligible for objects larger than @xmath64 km .
therefore larger objects , such as those required for the formation of the raditladi and rachmaninoff basins , are mainly produced by dynamical chaos loss @xcite .
those simulations show that the rate of large impactors decreased by a factor of 3 over the last 3 ga . another source of large impactors is the sporadic direct injection into strong resonances due to collisions @xcite .
the present neos average impact probability with mercury in the size range of 14 - 16 km , is of about one impact every 3.3 ga , in agreement with the proposed timescales of the formation of rachmaninoff and raditladi .
+ the authors wish to thank p. michel for helpful discussions on the cratering processes on a layered target .
we also wish to thank a. morbidelli for discussions regarding the neo population .
finally , we thank the referees ( c. chapman and an anonymous one ) for providing very interesting comments , that helped to improve our work .
blewett d.t . ,
robinson m.s . ,
denevi b.w . ,
gillis - davis j.j . , head j.w .
, solomon s.c .
, holsclaw g.m . , mcclintock w.e . , 2009 .
multispectral images of mercury from the first messenger flyby : analysis of global and regional color trends .
earth planet .
lett . , 285(3 - 4 ) , 272 - 282 .
bottke , w. f. , jedicke , r. , morbidelli , a. , petit , j .-
m . , & gladman , b. 2000 . understanding the distribution of near - earth asteroids .
science , 288 , 2190 - 2194 .
bottke , w. f. , morbidelli , a. , jedicke , r. , petit , j .-
m . , levison , h. f. , michel , p. , & metcalfe , t. s. 2002 . debiased orbital and absolute magnitude distribution of the near - earth objects .
icarus , 156 , 399 - 433 .
bottke , w. f. , vokrouhlick , d. , & nesvorn , d. 2007 .
an asteroid breakup 160 myr ago as the probable source of the k / t impactor .
nature , 449 , 48 - 53 .
collins , g. s. , melosh , h. j. , & marcus , r. a. 2005 .
earth impact effects program : a web - based computer program for calculating the regional environmental consequences of a meteoroid impact on earth .
, 40 , 817 - 840 .
collins , g. s. , kenkmann , t. , osinski , g. r. , wnnemann , k. 2008 .
mid - sized complex crater formation in mixed crystalline - sedimentary targets : insight from modeling and observation .
, 43 , 1955 - 1977 .
croft , s. k. 1985 .
the scaling of complex craters .
lunar planet .
conf . , 15 , 828 - 842 .
hawkins , s.e .
, boldt , j.d . ,
darlington , e.h . ,
espiritu , r. , gold , r.e . ,
gotwols , b. , grey , m.p . , hash , c.d . ,
hayes , j.r .
, jaskulek , s.e . ,
kardian , c.j . ,
keller , m.r .
, malaret , e.r . ,
murchie , s.l . ,
murphy , p.k . , peacock , k. , prockter ,
reiter , r.a .
, robinson , m.s . , schaefer , e.d . ,
shelton , r.g . , sterner , r.e . , taylor , h.w . , watters , t.r . , williams , b.d .
, 2007 . the mercury dual imaging system on the messenger spacecraft .
space sci .
rev . , 131 , 247 - 338 .
head j.w . ,
murchie s.l . ,
prockter l.m . ,
solomon s.c . ,
strom r.g . ,
chapman c.r . , watters t.r .
, blewett d.t .
, gillis - davis j.j . ,
fassett c.i .
, dickson j.l . , hurwitz d.m . ,
ostrach l.r . , 2009 .
evidence for intrusive activity on mercury from the first messenger flyby .
earth planet .
, 285(3 - 4 ) , 251 - 262 .
hiesinger , h. , head , j. w. , iii , wolf , u. , & neukum , g. 2001 .
new age determinations of lunar mare basalts in mare cognitum , mare nubium , oceanus procellarum , and other nearside mare .
lunar planet .
abstracts , 32 , 1815 .
holsapple , k. a. 1993 .
the scaling of impact processes in planetary sciences .
earth planet .
, 21 , 333 - 373 . , k. a. , housen , k. r. 2007 . a crater and its ejecta : an interpretation of deep impact .
icarus , 187 , 345 - 356 .
hrz , f. , grieve , r. , heiken , g. , spudis , p. , and binder , a. 1991 . lunar surface processes . in lunar sourcebook : a user s guide to the moon ( g. h. heiken , d. t. vaniman , and b. m. french , eds . ) cambridge university press , cambridge .
khan , a.k .
mosegaard , k.l .
rasmussen .
2000 . a new seismic velocity model for the moon from a monte carlo inversion of the apollo lunar seismic data .
j. geophys .
, 27(11 ) , 1591 - 1594 . ,
s. , morbidelli , a. , cremonese , g. 2005 .
flux of meteoroid impacts on mercury .
, 431 , 1123 - 1127 . ,
s. , mottola , s. , cremonese , g. , massironi , m. , martellato , e. 2009 . a new chronology for the moon and mercury
j. , 137 , 4936 - 4948 .
massironi , m. , cremonese , g. , marchi , s. , martellato , e. , mottola , s. , wagner , r.j .
mercury chronology revisited through mpf application on mariner 10 data : new geological implications .
j. geophys .
lett . , 36 , 21204 .
mcewen , a. s. , & bierhaus , e. b. 2006 .
the importance of secondary cratering to age constraints on planetary surfaces .
earth planet .
sci . , 34 , 535 - 567 .
mckinnon , w. b. , & schenk , p. m. 1985 .
ejecta blanket scaling on the moon and - inferences for projectile populations .
lunar planet .
abstract , 16 , 544 .
melosh , h. j. , 1989 .
impact cratering : a geologic process .
oxford university press , new york 1989 , pp .
minton , d. a. , & malhotra , r. 2010 .
dynamical erosion of the asteroid belt and implications for large impacts in the inner solar system .
icarus , 207 , 744 - 757 .
neukum , g. , & ivanov , b. a. 1994 , hazards due to comets and asteroids , 359 .
pike , r. j. 1988 .
geomorphology of impact craters on mercury .
mercury ( a89 - 43751 19 - 91 ) .
tucson , az , university of arizona press , 1988 , p. 165 - 273 .
prockter , l.m . ,
watters , t.r . ,
chapman , c.r .
, denevi , b.w . , head , j.w . ,
solomon , s.c . ,
murchie , s.l .
, barnouin - jha , o.s . ,
robinson , m.s . ,
blewett , d.t . , gillis - davis , j. 2009 .
the curious case of raditladi basin .
lunar planet .
abstract , 40 , 1758 .
prockter , l.m . ,
ernst , c.m . ,
denevi , b.w . ,
chapman , c.r .
, head iii , j.w . ,
fassett , c.i . ,
merline , w.j . ,
solomon , s.c . ,
watters , t.r . , blewett , d.t . ,
cremonese , g. , marchi , s. , massironi , m. , barnouin , o.s .
evidence for young volcanism on mercury from the third messenger flyby .
science , 329 , 668 - 671 .
quaide , w. l. , & oberbeck , v. r. 1968 .
thickness determinations of the lunar surface layer from lunar impact craters .
. res . , 73 , 5247 .
robinson m.s . ,
murchie s.l .
, blewett d.t . ,
domingue d.l .
, hawkins s.e .
, head j.w . , holsclaw g.m . , mcclintock w.e . ,
mccoy t.j . ,
mcnutt r.l . ,
prockter l.m . ,
solomon s.c .
, watters t.r .
reflectance and color variations on mercury : regolith processes and compositional heterogeneity .
science , 321 , 66 - 69 .
simmons g. , todd t. , wang h. 1973 .
the 25 km discontinuity : implications for lunar history .
science , 182 , 158 - 161 .
steffl , a. j. , cunningham , n. j. , durda , d. d. , & stern , s. a. 2009 , aas / division for planetary sciences meeting abstracts , 41 , # 43.01 stffler , d. , & ryder , g. 2001 .
stratigraphy and isotope ages of lunar geologic units : chronological standard for the inner solar system .
space sci .
, 96 , 9 - 54 .
strom , r.g . ,
chapman , c.r .
, merline , w.j . ,
solomon , s.c . ,
head , j.w .
mercury cratering record viewed from messenger s first flyby .
science , 321 , 79 .
tokosoz m.n .
, press f. , dainty a. , a. , anderson k. , latham g. , ewing m. dorman j. , lammlein d. , sutton g. , duennebeir f. 1972 .
structure composition and properties of lunar crust .
lunar planet .
conf . , 3 , 2527 - 2544 .
wagner , r. , head , j. w. , wolf , u. , & neukum , g. 2002 .
stratigraphic sequence and ages of volcanic units in the gruithuisen region of the moon .
j. geophys .
, 107 , 5104 zappala , v. , cellino , a. , di martino , m. , migliorini , f. , & paolicchi , p. 1997 .
maria s family : physical structure and possible implications for the origin of giant neas .
icarus , 129 , 1 - 20 . | in this paper we present a crater age determination of several terrains associated with the raditladi and rachmaninoff basins .
these basins were discovered during the first and third messenger flybys of mercury , respectively .
one of the most interesting features of both basins is their relatively fresh appearance .
the young age of both basins is confirmed by our analysis on the basis of age determination via crater chronology .
the derived rachmaninoff and raditladi basin model ages are about 3.6 ga and 1.1 ga , respectively .
moreover , we also constrain the age of the smooth plains within the basins floors .
this analysis shows that mercury had volcanic activity until recent time , possibly to about 1 ga or less .
we find that some of the crater size - frequency distributions investigated suggest the presence of a layered target .
therefore , within this work we address the importance of considering terrain parameters , as geo - mechanical properties and layering , into the process of age determination .
we also comment on the likelihood of the availability of impactors able to form basins with the sizes of rachmaninoff and raditladi in relatively recent times .
mercury , raditladi basin , rachmaninoff basin , craters , age determination |
radio pulsars are generally highly polarised sources , and the propagation of their radiation through the magneto - ionised interstellar medium ( ism ) is affected by faraday rotation .
measurement of the rotation measure ( rm ) , combined with the dispersion measure ( dm ) for each pulsar , provides a good estimate of the average magnetic - field strength in the direction of the line of sight to many of the known pulsars . with the high sensitivity and broad bandwidths of current observing instruments , it is possible to accurately measure rms within a single band , thus avoiding calibration issues of observing with different receivers in different bands
. furthermore , observing pulsars with high temporal resolution allows rm measurements at different pulse longitudes , provided the signal to noise ratio of the linearly polarised component is sufficient .
ramachandran et al .
( 2004 ) presented a phase - resolved rm analysis for psr b2016 + 28 . in an attempt to explain the variation in rm by @xmath2 rad m@xmath1 across the pulse profile
, they conducted a single - pulse analysis of the polarization .
they found that there are clearly two , almost orthogonal polarization modes across the pulse profile , with slightly different spectra ( as also postulated in karastergiou et al .
2005 and demonstrated in smits et al .
superposition of these modes results in a net change in the polarization position angle with frequency , leading to the measurement of a variable rm across the pulse .
although the effect measured in their study rotates the position angle ( pa ) with frequency , they concluded that it is not faraday rotation intrinsic to the pulsar magnetosphere .
the authors clearly demonstrate that for psr b2016 + 28 the position angle does not have a quadratic dependence on wavelength : i.e. @xmath3 .
a very recent work by @xcite showed that even small amounts of interstellar scattering can significantly deform the steep features of pa profiles ( e.g. orthogonal mode jumps ) , and can provide a simple explanation for the rm variations observed by ramachandran et al .
the latter is caused by the frequency dependence of the scattering constant , which causes a frequency - dependent amount of polarised intensity to be transfered from a given phase bin to later phases .
regions where the pa profile has a steep gradient are worst affected , since it is at those phases where the scattered portion of the polarised intensity has the most impact on the shape of the pa profile : flat profiles can only stay flat , but the kinks and wiggles in pa profiles will be distorted by the redistribution of the polarised intensity . hence , karastergiou noted that the polarization of core components , where pa profiles are steepest , is affected the most .
the author concluded that scattering can smear orthogonal polarization modes so that they appear slightly less `` orthogonal '' .
in this paper we concentrate on pulsars for which we have recently published rm values ( e.g. noutsos et al .
our aim is not only to fit for pulse phase resolved rms , as was done in ramachandran et al . , but to closely investigate whether the polarization position angles obey @xmath0 fits .
we investigate the possible causes for this particular pa frequency dependence and discuss implications on rm measurements performed using different techniques .
the data used in our analysis were collected with the parkes 64-m radio telescope during two observing periods : in 2004 , august 31 to septemeber 1 , and in 2006 , august 24 to 27 .
we observed at the frequency band of the 20 cm h - oh receiver , installed at the focus cabin of the parkes telescope .
the receiver is equipped with a pair of orthogonal , linear feeds that are sensitive to polarised emission with an equivalent system flux density of 43 jy at 20 cm ( johnston 2002 ) .
the 2004 data were observed with a central frequency of 1.375 ghz and 256 mhz bandwidth , while in 2006 we observed with a central frequency of 1.369 ghz and 256 mhz bandwidth .
the wide - band correlator ( wbc ) back end that was used in both observations is capable of splitting the available bandwidth into 1,024 frequency channels , on each of four polarizations , @xmath4 , @xmath5 , @xmath6 and @xmath7 , and producing 10-s de - dispersed pulse subintegrations with up to 2,048 phase bins across the pulse profile . in noutsos
( 2008 ) , we outline the method used to derive a single rm value for each pulsar we observed . here
, we performed the same analysis of the polarised emission in each longitude bin , across the total pulse width .
our aim was to produce rm profiles as a function of longitude for each observed pulsar .
prior to our analysis , the raw data from the wbc were subjected to the necessary process of rfi mitigation and polarization calibration .
we processed the calibrated data using the psrchive software package @xcite .
it is expected that the plane of linearly polarised emission will be rotated by the magnetised ism it traverses on its way to the earth : this is the well - known interstellar faraday effect .
the amount of rotation , @xmath8 , for a given pulsar should be constant for a specific frequency but scales as @xmath9 across the observing band . by definition ,
the rotation between frequency @xmath10 and infinite frequency gives the pa at frequency @xmath10 : @xmath11 since the rm is a physical property of the ism , it is expected to be constant for a given ism configuration between the pulsar and the earth .
its value depends on the magnitude and direction of the magnetic field , @xmath12 , along the propagation path of the polarised emission , @xmath13 ; the free - electron density , @xmath14 , along @xmath13 , and the length of @xmath13 ; it is given by the path integral @xmath15 where @xmath16 is directed from the pulsar to the earth .
apart from the pulsar distance , none of the remaining physical quantities , on which rm depends , are directly measurable .
however , given a finite observing bandwidth , one can derive rm by fitting the pa across the band with the quadratic function of equation [ eq : paeq2 ] .
the pa of each bin @xmath17 in each frequency channel @xmath18 was calculated as @xmath19 .
having previously developed the software capable of such quadratic fits ( see noutsos et al .
2008 for details ) , we performed bin - wise fitting across the longitude range inside the pulse and derived a value of rm for each individual longitude bin . in order to increase the s / n per frequency channel , @xmath18 , before fitting across the whole available bandwidth we averaged the frequency channels by a factor 16 . for a single averaging factor , as is the case in this analysis , only the statistical error on each rm value can be calculated , and the fitting routine gives the best integer rm for each bin .
finally , we plotted pulse profiles of the total flux density , the linear- and circular - polarization components , and the pa and rm profiles across the pulse ; we only plotted significant pa values with @xmath20 and rm values with @xmath21 rad m@xmath1 .
the resulting 19 rm profiles are shown in fig . [
fig : largermvar1 ] and fig .
[ fig : smallrmvar1 ] .
9 pulsars show large rm variation across the pulse and 10 pulsars show a rather constant rm profile . in these plots ,
the bottom panel shows the polarization profile , where solid lines denote total power ; dashed lines , linear polarization ; and dotted lines , circular polarization .
the middle panels show the integrated pa profile and the top panels , the rm profile . in the interest of characterisation
, we have included a number of properties of the pulsar emission and the ism in the direction of the 19 pulsars , in table [ tab : pulprop ] .
* psr j0835@xmath224510 ( 2004 & 2006 data ) . *
the vela pulsar has a highly polarised , yet relatively simple pulse profile in both the 2004 and 2006 data .
both profiles appear scattered .
the pa shows the characteristic swing which has been used as a strong argument for the rotating vector model ( radhakrishnan & cooke 1969 ) .
the most striking profile of vela is its rm profile , which shows a clear swing across the pulse that resembles a sideways ` s ' ( peak - to - peak variation @xmath23 rad m@xmath1 ) .
the shape of the rm profile persists through both 2004 and 2006 data sets , albeit the earlier of the two profiles displays a somewhat higher peak - to - peak variation of the rm ( @xmath24 rad m@xmath1 ) .
* psr j0942@xmath225552 ( 2004 data ) . *
the pulse profile of this pulsar has 3 distinct components , with the peaks of linear polarization being slightly offset from the corresponding ones in the total power .
the pa profile shows an orthogonal jump at the bridge emission between the leading and central components and at the exact phase where @xmath25 vanishes .
the rm profile shows a steep decline of high - s / n points coincident with the leading end of the central component and an equally steep increase , coincident with the leading component ; this behaviour occurs around the orthogonal pa jump . in both cases , the rm changes by at least 20 rad m@xmath1 . at the central peak
s maximum the rm fluctuates within a few rad m@xmath1 .
* psr j1056@xmath226258 ( 2004 data ) . *
the total - power profile of this pulsar appears very slightly stretched toward later pulse phases .
it has nearly zero circular polarization but @xmath26% linear .
the pa profile sweeps smoothly across the pulse , without showing any breaks .
the rm profile is roughly consistent with the published rm around the the flux maximum , but it drops suddenly near the pulse edges , giving it an overall shape of a swing across the entire pulse .
@xcite performed a comparison between the 20 cm and 10 cm profiles of this pulsar and concluded that the linear polarization , and hence the pa profile , must undergo significant frequency evolution .
they supported their argument by deriving a pulse - averaged rm value for psr b1056@xmath226258 that was much closer to the published one ( 6 rad m@xmath1 ) , whereas if no evolution was assumed that value was @xmath27 rad m@xmath1 .
however , this pulsar shows a peak - to - peak rm variation across the pulse of almost 100 rad m@xmath1 the highest in our sample which dwarfs any possible rm error due to profile evolution .
* psr j1243@xmath226423 ( 2004 data ) . *
the central pulse of this pulsar s profile consists of two gaussian components , whereas a third , much weaker component is present at the far leading edge of the profile .
the circular polarization profile resembles an unsusual double swing , with its minimum value being roughly coincident with the maximum @xmath25 .
there is an orthogonal pa jump after the leading component .
the pa profile corresponding to the central pulse components displays a full ` s'-shaped swing , characterised by wiggles often seen in pa profiles .
the rm profile of this pulsar is extremely complex : it shows a clear structure that departs from flatness , with a leading swing followed by a sharp peak and a steep decline of the rm .
the peak - to - peak variation of the profile is @xmath28 rad m@xmath1 .
the maximum rm ( @xmath29 rad m@xmath1 ) appears coincident with the inflexion point of the pa swing .
* psr j1326@xmath225859 ( 2004 & 2006 data ) .
* this pulsar s profile is the sum of at least two components : a weak leading one and a strong central one that leads into a trailing tail , perhaps indicative of a third , very weak component .
the linear polarization of the central component is moderatly high , with more than 30% of its total flux being polarised , whereas the leading component is just over 20% polarised .
even more interesting is the evident swing of the circular polarization under the central component . like in the case of psr j1157@xmath226224
, this may well be hinting at emission from near the magnetic axis . in both epochs ,
the pa profile is broken by an orthogonal jump near the peak of the leading component , at the exact point where @xmath25 vanishes . at later phases , the pas form an unusual swing interrupted mid - way by a bulge that is coincident with the maximum of the linear polarization .
the rm profile clearly shows a smooth functional dependence on phase that deviates from flatness ; like in the case of vela , this dependence is exactly reproduced in both 2004 and 2006 data , which argues against this being an instrumental effect .
the form of the rm function resembles a double swing with 2 local minima separated by a local maximum .
the trailing of the two minima is exactly coincident with the peak in total power .
* psr j1453@xmath226413 ( 2004 data ) . *
the pulse profile of this pulsar is a composition of a relatively weak leading component and a much stronger central one .
however , in the linear - polarization profile these two components are of comparable intensity .
the pa profile swings along the entire central component but is disrupted by an orthogonal jump that coincides with the merging point of the two components linear polarization , where @xmath30 .
the rms across the pulse show a moderate amount of scattering that prevents us from being certain about the existence of a possible structure .
on one hand , there is a clear break in the rms due to the jump in the pas . on the other hand , even away from the jump there seems to be a dip in the rms that is coincident with the peak of the central component s linear polarization .
* psr j1644@xmath224559 ( 2004 data ) . *
the total power of this pulsar can be described by a single gaussian with an extended , trailing scattering tail ( see also analysis by johnston 2004 ) . a hint at a second , very weak component can be seen near the leading edge of the main pulse .
the maximum of the linear polarization occurs at slightly earlier phase than that of the total - power maximum .
in addition , this pulsar exhibits a swing in the circularly polarised profile .
the pa profile shows a more complex picture , in comparison : it features two smooth bumps across the main pulse , which are only interrupted by an approximately orthogonal jump , coincident with the bridge emission between the main pulse and the weak leading component ; it also shows the obvious signs of scattering on polarization shown in li & han ( 2001 ) . the rm profile features a clear structure that comprises at least three local maxima and an equal number of local minima ; none of these extremes is exactly coincident with either the total flux or the linearly polarised flux extremes .
the fluctuation of the rm across the pulse appears mildly correlated to the pa profile .
* psr j1745@xmath223040 ( 2004 data ) .
* there are three , distinct , highly polarised components in this pulsar s profile .
@xcite performed multifrequency observations of this pulsar and noted the significant frequency evolution of its profile as well as the apparent increase of the linear polarization fraction with frequency .
the pa profile resembles an upside - down @xmath31 across the trailing and central components , which is interrupted by an orthogonal jump occuring between the trailing components and the leading one .
the rm profile shows a seemingly constant downwards slope for the most part of the two trailing components but kinks downwards at the leading edge of the central component . above the leading component ,
the rms are scattered , although an upwards slope can perhaps be inferred from the 6 data points coincident with this component . * psr j2048@xmath221616 ( 2006 data ) . *
three gaussian components can be discerned in the total profile of this pulsar .
the linear polarization is strong in the leading and trailing of the three components but relatively weak in the central one .
the pa profile resembles an ` s'-shaped swing that rotates the pas smoothly across the profile ( see also johnston et al . 2007 ; johnston et al .
the rm is rather constant across the leading component but dips by @xmath32 rad m@xmath1 at the central , least - polarised component .
a sudden rise of the rms is observed on the trailing edge of the trailing component , although the s / n of the linear polarization is already low in that phase region . * psr j0134@xmath222937 ( 2006 data ) . *
the pa profile of this pulsar is flat everywhere , apart from the phase region around the pulse maximum , where the profile is broken by an orthogonal jump .
the jump is also coincident with the narrowest of the two distinct components present in the linearly polarised flux , the other being broad and coincident with the long leading tail of the total flux .
the rm profile of this pulsar shares the same flatness , which seems to be disturbed by the presence of the orthogonal jump .
however , the rm fluctuation occurs only over a few phase bins , and its magnitude is of the same order as at least three other instances away from the orthogonal jump .
* psr j0738@xmath224042 ( 2004 & 2006 data ) . * the polarization of this pulsar was shown at two frequencies ( 1.4 and 3 ghz ) in karastergiou & johnston ( 2006 ) .
the position angle shows two orthogonal - mode jumps . the rm profile ,
although generally flat , is below the average rm value in the leading part of the profile ; it then turns above the average rm value for the rest of the profile , after the first orthogonal jump . in the dual frequency data of karastergiou & johnston , the frequency evolution of the position - angle profile is in agreement with what is seen here :
the first segment of the pa profile has a different frequency dependence to the rest of the profile .
* psr j0742@xmath222822 ( 2004 & 2006 data ) .
* this pulsar comprises a complex , 7-component profile ( see e.g. kramer 1994 ; karastergiou et al .
2005 ) . despite the complexity of the pulse itself , the pa and rm profiles display a relatively simple dependence on phase , with the former being a simple sweep across the profile and the latter being rather flat .
it is perhaps noteworthy that a comparison between the profiles of 2004 and 2006 shows that the average rm in the earlier observations appears slightly higher than the published value , whereas the opposite is observed in 2006 .
the @xmath33 difference between the two data sets is @xmath34 rad m@xmath1 . * psr j0907@xmath225157 ( 2006 data ) . *
the total - power and linear - polarization profiles of this pulsar have two components .
the pa profile is a smooth , continuous sweep , without any breaks , and one can notice that it slopes slightly upwards , toward the trailing edge of the pulse profile .
the rm profile is flat and consistent within the errors with the published rm . * psr j1057@xmath225226 ( 2006 data ) . *
this young pulsar ( @xmath35 y ) is an interesting and a peculiar one , and one of a few that are visible at all wavelengths ( see e.g. vaughan & large 1972 ; cheng & helfand 1983 ; migniani et al .
1997 ; fichtel et al .
even more interesting is its strong radio interpulse , which is separated from the main pulse by roughly 150@xmath36 and has the same energy as the main pulse over a wide range of radio frequencies ( mcculloch et al .
1976 ; weltevrede & wright 2008 _ mnras submitted _ ) ; it has been suggested that we observe emission from both magnetic poles ( lyne & manchester 1988 ; wang et al .
furthermore , the steepness of the pa swing for the main pulse and the interpulse differs significantly , which has given rise to the suggestion that the corresponding emission is generated at different heights in the magnetosphere ( wang et al .
2006 ) . due to the higher degree of linear polarization ( @xmath37 ) and the modest s / n of the pulse in the 2006 data
, we chose to study only the main pulse for rm variations .
the pa profile of the main pulse is flat , with no breaks , and shows rotation of no more than 60@xmath38 across the entire pulse .
the rm profile is also consistent with a constant rm .
* psr j1157@xmath226224 ( 2004 data ) .
* this pulsar shows a simple total - power profile , with substantial amounts of linear ( @xmath39% ) and circular polarization .
the pa profile exhibits a smooth but complex swing , broken only by an orthogonal jump half - way along the trailing edge of the pulse .
the rm profile also swings , although the edges are too noisy and the magnitude of the effect too small to draw a certain conclusion .
* psr j1253@xmath225820 ( 2006 data ) .
* this pulsar s pulse profile can be fitted with a single gaussian component , and the same pulse shape is shared by the linearly polarised flux , which accounts for a large fraction of the total observable emission .
the pa shows a smooth variation throughout .
the rm profile shows little deviation from flatness , mainly at the leading edge .
* psr j1359@xmath226038 ( 2004 data ) .
* this pulsar s profile is @xmath40% linearly polarised and @xmath41% circularly polarised .
the pa profile is rather flat across the pulse .
the rm profile is consistent with a constant value for the most part of the profile , diverging slightly at the tail end , similarly to psr j1056@xmath226258 .
however , at that point the s / n is greatly reduced so solid conclusions can not be extracted . * psr j1705@xmath221906 ( 2006 data ) .
* this is another highly polarised pulsar with a main pulse composed of multiple components : despite having been classified as a three - component pulse , at least five components can be discerned in both the total power and the linear polarization profiles in our data ( rankin 1990 ; gould & lyne 1998 ) .
in addition , a much simpler , single - peaked interpulse is also present in this pulsar s profile , which is separated by roughly @xmath42 from the main pulse . at 20 cm ,
the interpulse is almost 100% linearly polarised , which has also been noted in previous work ( see e.g. karastergiou & johnston 2004 ) . in this work
, we chose to examine only the main pulse for rm variations , due to its larger width and higher s / n . the pa profile is a gradual sweep across the entire main pulse .
the rm profile appears flat , on average , although a coherent jitter over a - few - bin scale is evident .
the only possible departure from flatness appears to coincide with the leading outrider .
* psr j1807@xmath220847 ( 2004 data ) . *
another case of a multi - component , total - power profile , with low linear polarization and with a central component surrounded by at least three outriders .
the pa profile is broken twice by orthogonal jumps , which coincide with the minima of the linear polarization ; this has also been noted in a separate analysis by @xcite .
the rm profile is quite noisy , owing to the low linear polarization , but flat , on average , across the entire pulse . in order to check for the presence of any frequency - dependent , systematic effect that could cause a change of the pa across the frequency band that is different to the @xmath0 law of faraday rotation
, we examined in more detail the fits that produced the calculated rms for vela , j1243@xmath226423 and j1326@xmath225859 .
[ fig:0835rmfits ] , [ fig:1243rmfits ] and [ fig:1326rmfits ] show the rm fits for the minimum- and maximum - rm bins ( denoted by arrows ) that reside near the central part of the profile of the three aforementioned pulsars .
it is important to note that the associated errors on the fitted rms , being only statistical , would be underestimated if there was an additional , systematic , non - faraday component that affects the pas . in order to reveal the presence and magnitude of a possible systematic discrepancy between the fitted function and the data , for each fit we also plotted the fit residuals and their distribution ( shown below the fit plots , in fig .
[ fig:0835rmfits ] , [ fig:1243rmfits ] and [ fig:1326rmfits ] ) .
the quality of the data and the goodness of the fit are evident in these figures . upon closer inspection
, however , the residual pas of all three pulsars , but mostly j1243@xmath226423 and j1326@xmath225859 , show evidence of some weak , sinusoidal fluctuations around the mean .
it is difficult to attribute some significance to this , especially given the gaussian - like distributions of the residuals .
the conclusion therefore is that , if there is indeed a physical mechanism that affects the pa rotation , apart from the interstellar faraday effect , we are unable to distinguish it from the dominant @xmath9 law .
we also investigated the possibility that the rm variations observed are caused by erroneous estimation of the dm , with , say , a @xmath43 offset from the true value .
such an offset is translated into a dispersive delay between the pulse arrival time at a reference frequency @xmath44 ( usually the centre frequency of the band ) and that at frequency @xmath45 , where @xmath46 . in that case , the delay in ms is given by @xmath47 which also corresponds to a phase offset in the pulse window , @xmath48 , with @xmath49 the pulsar period .
if this phase offset is not accounted for , the calculation of the pa in each frequency channel will correspond to a pulse longitude that is shifted with respect to that at the reference frequency ; this may introduce a frequency - dependent shift in the pa of any given phase bin . in order to eliminate this as the cause of the observed rm variations ,
we de - dispersed our data using the published dm value , as well as two other values that were @xmath50 away , either side of the published one , and generated again the rm profile in each case .
the published dms for our pulsars are shown in table [ tab : pulprop ] . by checking the resulting rm profiles for all pulsars , we concluded that there is no way of recovering a flat rm curve for any reasonable departure from the published dm .
we therefore rule out dm errors as a possible cause of rm variation .
our analysis reveals that the rm profiles of pulsars with large rm variations are diverse : they can be as simple as the ` s'-shaped profiles of vela and psr j1056@xmath226258 or as complex as those of psr j1243@xmath226423 and psr j1644@xmath224559 .
the peak - to - peak variation of those pulsars rms across their profiles was found to be as little as a few rad m@xmath1 ( e.g. vela in 2006 ) but also as large as 100 rad m@xmath1 , for psr j1056@xmath226258 . despite the hints of correlation between individual features in the flux and pa profiles and those in the rm profiles ( see e.g. psr j1326@xmath225859 ) , in most cases these correlations seem coincidental . however , a general observation is that the greatest rm fluctuations across the pulse profile seem coincident with the steepest gradients of the pa profile , whereas pulsars with flat pa profiles show little or none at all rm variation .
good examples that support this argument are psr j1243@xmath226423 and psr j2048@xmath221616 , whose rm departure from the mean is maximised around the inflexion point of the pa swing ; in contrast , psr j0907@xmath225157 and psr j1057@xmath225226 , having flat pa profiles , show insignificant rm variations .
the subset of pulsars with low rm variations adheres to the theoretical predictions .
more than half of those pulsars have phase - resolved rms that are scattered around their respective published rm value , thus showing convincingly longitude - independent rm values across the pulse .
however , there are cases like those of psr j1157@xmath226224 , psr j1253@xmath225820 , psr j1359@xmath226038 and psr j1705@xmath221906 , where the rm scatter across the pulse is clearly not gaussian : these profiles show a clear structure but with a small rms spread .
an interesting case is that of psr j0742@xmath222822 , where although no significant rm variations exist across the profile , the longitude - averaged rm across the pulse is higher in 2004 data than in 2006 .
nevertheless , one has to be careful when drawing the above conclusions ; the main statement is that any rm variations with pulse phase are smaller than the statistical error of the measurement .
we have demonstrated that the rm variations observed are not due to measurement errors , and that the pa rotation in all cases follows the @xmath0 law of faraday rotation . in the following we consider three possible explanations : ( a ) faraday rotation within the pulsar magnetosphere , ( b ) the superposition and frequency dependence of quasi - orthogonal polarization modes , and ( c ) interstellar scattering and its effects on the polarization .
faraday rotation through the ionised ism has the effect of rotating the pa with frequency . our data
, however , comprise of full sets of the four stokes parameters , @xmath4 , @xmath5 , @xmath6 and @xmath7 , so we can observe the frequency dependence of the entire polarization vector ( @xmath51 ) as a function of frequency , for specific high - s / n phase bins .
an example for one phase bin of psr j1326@xmath225859 is shown in fig .
[ fig : poincare ] , where the tip of the poincar vector is plotted across 32 frequency channels .
this plot demonstrates that , together with the pronounced effect of pa rotation ( around the vertical axis ) , a small and rather systematic change in the fraction of circular polarization can be seen .
we obtained similar plots for four additional pulsars in our sample that displayed large rm variations : namely vela , psr j1243@xmath226423 , psr j1644@xmath224559 and psr j2048@xmath221616 . in nearly all cases
, we found that the trace of @xmath52 on the poincar sphere does not strictly follow paths of constant latitude , as in fig .
[ fig : poincare ] . the wave modes in the cold and ionised ism
are purely circularly polarised .
therefore , the phase offset introduced by propagation through this medium results in the well - documented changes of the pa . on the contrary , in a medium of highly relativistic charged particles , the natural modes are close to linearly polarised ( sazonov 1969 ; melrose 1997a ) . as a consequence ,
different phase offsets between the modes will result in a conversion between linear and circular polarization . in a pulsar s magnetosphere ,
the observed degrees of circular polarization indicate that the natural wave modes are elliptically polarised , which suggests that propagation through this medium will generate some frequency - dependent conversion between linear and circular polarization .
an extensive discussion on this can be found in @xcite .
if such a generalised faraday rotation is taking place within the pulsar magnetosphere and giving rise to rm variations as well as changing the degree of circular polarization in the observed frequency band , we expect the greatest change in @xmath7 to coincide with the greatest variations in rm . in order to test this , we first calculated the fraction of circularly polarised intensity , @xmath53 , for each frequency channel across the band .
the division by @xmath4 removes any fluctuations of @xmath7 caused by scintillation through the interstellar medium , whereas any change in @xmath7 that originates in the magnetoshpere should remain unaffected .
subsequently , we fitted a linear function , @xmath54 , to the data and calculated the change in @xmath53 across the band , based on that function .
this was done to get a good estimate of the total effect while minimizing the influence of instrumental noise .
we found linear fits to describe the data sufficiently well .
the resulting profiles of @xmath55 as a function of longitude , for vela , psr j1243@xmath226423 , psr j1326@xmath225859 , psr j1644@xmath224559 and psr j2048@xmath221616 are shown in fig .
[ fig : vdvprofiles ] . a side - by - side comparison of the rm and @xmath55 profiles , shown in fig .
[ fig : vdvprofiles ] , makes clear that none of the pairs of profiles seems correlated .
there is certainly little evidence that the maximum rm deviation from the mean and the maximum change in circular polarization across the band occur at the same pulse phase .
therefore , to first order , the assumption that the observed rm variations are caused by magnetospheric faraday rotation is deemed unlikely .
another factor that could potentially affect the measured rm is the _
incoherent _ superposition of quasi - orthogonal modes of polarization ( see e.g. ramachandran et al .
the superposition of strictly orthogonal modes results in an observed pa of the strongest of the two modes , regardless of the ratio of the intensity of the modes ( e.g. backer et al 1976 ; cordes et al .
the reason is that orthogonal modes are represented by anti - parallel vectors on the poincare sphere ; regardless of their relative lengths , the addition is always pointing in the direction of either of the two modes .
this is not the case for non - orthogonal modes : in this case , the sum of the mode vectors points at an angle that depends on the degree of non - orthogonality and the relative intensity of the modes . as a consequence , if the ratio of the mode intensities changes as a function of frequency , the observed pa will also rotate .
the question is under what conditions this could resemble faraday rotation . in the pulsars where significant rm variations are seen ,
the variations often exceed 20 rad m@xmath1 across the pulse profile . at the frequency and bandwidth of these observations
, this corresponds to a variation in pa by @xmath56 .
therefore , in order for quasi - orthogonal polarization modes to present a possible explanation for the observed rm variations , they should be capable of inducing such a pa rotation . from polarization studies of 18 pulsars in the european pulsar network ( epn ) archive ,
@xcite conclude that the ratio of the average intensity of the orthogonal modes for those pulsars follows a power law with spectral index that does not typically exceed @xmath57 .
assuming linear modes ( @xmath58 and @xmath59 ; the degree of circular polarization is generally very low ) , then at a reference frequency @xmath44 the mode intensity ratio is @xmath60 , and the form of this power law is @xmath61 for given values of @xmath62 , we can compute @xmath63 across the frequency band we are interested in : in this case , between 1300 and 1556 mhz .
we have numerically computed the total rotation of the pa across the band ( hereafter @xmath64pa ) , for different values of @xmath62 and different angles of non - orthogonality , @xmath65 .
the latter angle is defined on the poincar sphere as the supplementary to the angle between the two modes .
figure [ fig : dpafunc ] shows @xmath64pa versus @xmath65 , for various values of @xmath62 , which equate to certain degrees of linear polarization .
the plot on the left shows examples where @xmath66 as opposed to @xmath67 for the plot on the right .
both plots show a set of lines corresponding to the same degrees of polarization , @xmath25% , which is approximately @xmath68 .
two things are immediately evident : first , the greatest rotation of pa happens for small values of @xmath65 ; second , and most important , the only possibility that @xmath64pa@xmath69 occurs when the fractional polarization remains under 10% and only in the specific case where @xmath70 , i.e. the case where the mode with the steepest spectral index is brightest at the lowest frequency @xmath44 . in our data , there are clear examples of high @xmath64rm and high @xmath25% , the most prominent being vela and j1326@xmath225759 .
it is therefore unlikely that non - orthogonal polarization modes provide a general explanation for the rm variations in the data .
as was mentioned earlier in this paper , some of the pulsars we studied show signs of interstellar scattering in their flux profiles : vivid examples are psr j1056@xmath226258 , justified by its high dm of 320.3 pc @xmath71 , and psr j1644@xmath224559 , with a dm of 478.8 pc @xmath71 the highest in our sample .
we checked for a possible correlation between the amount of interstellar scattering and the magnitude of the observed rm variations .
the effect of interstellar scattering is usually quantified by the scattering timescale , @xmath72 . assuming a kolmogorov spectrum for the density fluctuations along the sightline to the pulsar , @xmath72 has a roughly quadratic relation to dm for low - dm pulsars ( i.e. @xmath73 ) but can be much steeper for high - dm pulsars : i.e. @xmath74 ( e.g. sutton 1971 ; ricket 1977 ) .
more recently , based on a global fit to 98 scattering times of high - dm pulsars , @xcite showed that a parabolic curve of the form @xmath75 where @xmath44 is the observing frequency , is a reasonable description of the trend of @xmath72 as a function of dm .
hence , as long as we are refering to the same observing frequency , and since there are only a few available measurements of @xmath72 for our pulsar sample ( see table [ tab : pulprop ] ) , the dm is a convenient alternative to the amount of scattering a pulsar s emission is subject to .
we have quantified the magnitude of the rm variations for those pulsars whose rm profiles show signs of a longitude - dependent structure , in terms of the peak - to - peak distance of their rm profile , @xmath76 .
[ fig : drmmaxdm ] shows a scatter plot of @xmath76 against dm for 9 pulsars .
although there is perhaps no clear correlation , it is striking that the pulsars that typify high rm variations are also high - dm pulsars .
this includes the most prominent examples of rm variation with pulse phase , such as psr j1243@xmath226423 , psr j1326@xmath225859 and psr j1644@xmath224559 . in a recent paper
, @xcite showed that small amounts of scattering and orthogonal polarization modes of emission in the pulsar magnetosphere can result in a frequency dependent pa that can be falsely attributed to faraday rotation .
this explanation is extremely simple and may well be the key to understanding the high - dm examples .
it is also strengthened by the observation that most of the low - dm pulsars show relatively smooth pa profiles , devoid of the bumps and wiggles seen in the higher dm profiles . the overall correlation in fig .
[ fig : drmmaxdm ] is perhaps reduced by the fact that rm measurements can only be made for pulse phases with significant linear polarization .
we have shown phase - resolved rm profiles for 19 pulsars , 9 of which show substantial rm variations across the pulse profile .
these variations appear in most cases to show coherent rather than noisy behaviour .
the high s / n of our data and our robust fitting technique have shown that whatever the cause of the pa rotation with frequency , it is inseparable from the @xmath9 law of interstellar faraday rotation within our observed band . we have considered a number of possibilities that may result in the observed phenomena . for example , we have ruled out the possibility that de - dispersion with a slightly erroneous dm may be responsible for the data here . we have discussed the possibility that the rm variations are of magnetospheric origin .
two cases were examined , that of generalised faraday rotation due to the relativistic plasma in the pulsar magnetospheres , and the existence and superposition of quasi - orthogonal modes of polarization . in the case of generalised faraday rotation , we examined whether pulse phases with the highest rotation of the pa also showed the highest change of the fractional circular polarization .
our first - order investigation turned out negative results , although it is true that the theory for generalised faraday rotation is not complete and does not provide us with good handles on the rate at which polarization rotations occur with frequency . as concerns quasi - orthogonal modes , we find that it is impossible to obtain a large rotation of the pa within our observing frequency band and observe large degrees of linear polarization at the same time .
our analysis indicates that there are specific cases of the data that can be explained by this process , but it certainly does not form a general framework for explaining rm variations across the pulse . finally , this work supports the ideas presented in @xcite , that interstellar scattering can distort the polarization in a frequency - dependent way .
this may resemble faraday rotation and will generally be pulse - phase dependent .
it is important to emphasise that the possibility of magnetospheric faraday rotation is not totally eliminated by this work ; we merely conclude that it is an unlikely explanation for rm variations of the observed magnitude .
there is a possibility that both the magnetospheric mechanisms and interstellar scattering are responsible for the observed rm variations , along with other possibe mechanisms which we have not considered . whichever the explanation , certain care must be taken during rm - determination experiments from pulsars to correctly identify the amount of pa rotation that is due to the iono - magnetised interstellar medium . especially for the studies of the planar component of the galactic magnetic field , where pulsars with high dms and rms are used , the identification of the ism - induced faraday rotation and the correct calculation of the associated rm plays an important role in our efforts to model the interstellar magnetic fields .
hence , in light of the rm profiles presented herein , pulsar - rm determination from specific , high - s / n pulse bins can lead to considerable errors .
the australia telescope is funded by the commonwealth of australia for operation as a national facility managed by the csiro .
we would like to thank andrew lyne , don melrose and roy smits for useful discussions and comments on this manuscript .
.[tab : pulprop ] tabulated properties of the 19 pulsars we analysed in this work . columns 7 and 8 show the pulse - averaged fraction of the linearly and circularly polarised intensity , respectively .
the last column shows the magnitude of the interstellar pulse broadening as a percentage of the pulse s fwhm ; this is only known for 6 pulsars in our sample .
[ cols= " < , > , > , > , > , > , > , > , > , > , > " , ] profiles of the rm as a function of pulse longitude ( top panels ) , for 9 pulsars , observed at 20 cm with the parkes telescope , that exhibit large rm variations across their pulse profile .
the respective year of observation is shown at the top - left corner of each pulsar s rm profile . for two pulsars , namely the vela pulsar and psr j1326@xmath225859
, we had data from two observing sessions , in 2004 and 2006 .
the rm for each longitude bin was calculated from a quadratic fit to the pas of that bin across the frequency band .
only rm values with @xmath21 rad m@xmath1 are shown .
the horizontal , gray line across the rm profiles shows the published rm value for each pulsar .
the middle panels show the phase - resolved pa profiles , calculated from the integrated stokes profiles , @xmath5 and @xmath6 , across the respective frequency band . for each longitude bin ,
the pa was calculated as @xmath77 .
only pas corresponding to linear polarisation with @xmath78 are shown .
the bottom panels show the profile of the total flux density ( solid lines ) and the linear ( dashed lines ) and circular ( dotted lines ) polarization flux for each pulsar.,scaledwidth=80.0% ] profiles of the rm as a function of pulse longitude ( top panels ) , for 10 pulsars , observed at 20 cm with the parkes telescope , that exhibit small or no rm variations across their pulse profile .
the respective year of observation is shown at the top - left corner of each pulsar s rm profile . for two pulsars , namely psr j0738@xmath224042 and psr j0742@xmath222822 , we had data from two observing sessions , in 2004 and 2006 .
see fig .
[ fig : largermvar1 ] for more details.,scaledwidth=85.0% ] rm fits to the pas corresponding to the maximum- and minimum - rm values in the central part of vela pulsar s rm profile .
the set of plots to the left of the pulsar s j2000 name ( j0835@xmath224510 ) correspond to the 2006 data ; the ones to right of the legend correspond to data collected in 2004 .
the top plots in each set show the rm variation across the pulse ; below them , we show the rm fits corresponding to the minimum ( left ) and maximum rm ( right ) in the rm profile .
the minimum- and maximum - rm values selected for the fits are indicated with black arrows in the respective rm profiles . below each fit
we show the fit residuals in degrees , and at the very bottom the distribution of those residuals is shown.,scaledwidth=98.0% ] the rotation of the poincar vector , @xmath52 , as a function of frequency , across the frequency band , for psr j1326@xmath225859
. the 3d representation of the rotation is zoomed into the latitude range which contains the data .
the data points were best fitted with a circle of inclination @xmath79 ( solid black line ) , whereas a flat circle of zero inclination ( dashed black line ) gave a worse fit.,scaledwidth=55.0% ] top panels : the change in the fraction of circularly polarised intensity across the band as a function of pulse longitude , for vela , psr j1243@xmath226423 , psr j1326@xmath225859 , psr j1644@xmath224559 and psr j2048@xmath221616 . for comparison ,
we have included , in the middle and bottom panels , the rm and polarization profiles of each pulsar.,scaledwidth=100.0% ] the total pa difference between 1300 and 1556 mhz due to the frequency evolution of the relative strength of quasi - orthogonal modes , as a function of the angle of departure from orthogonality , @xmath65 , and for ( a ) increasing and ( b ) decreasing values of the initial relative intensity of the modes @xmath62 , as indicated.,scaledwidth=47.0% ] scatter plot of the maximum , peak - to - peak rm difference ( @xmath76 ) as a function of the pulsar dm , measured across the profiles of the 9 pulsars that show large rm variations and psr j1359@xmath80 , for which a lesser degree of rm variation is visible .
the arrow at the top of the frame indicates that the @xmath76 value measured for psr j1056@xmath226258 ( @xmath81 rad m@xmath1 ) is off the plot s scale .
the symbol size in this plot is proportional to the logarithm of the average flux at 1.4 ghz , @xmath82 .
, scaledwidth=47.0% ] | we have detected significant rotation measure variations for 9 bright pulsars , as a function of pulse longitude .
an additional sample of 10 pulsars showed a rather constant rm with phase , yet a small degree of rm fluctuation is visible in at least 3 of those cases . in all cases
, we have found that the rotation of the polarization position angle across our 1.4 ghz observing band is consistent with the @xmath0 law of interstellar faraday rotation .
we provide for the first time convincing evidence that rm variations across the pulse are largely due to interstellar scattering , although we can not exclude that magnetospheric faraday rotation may still have a minor contribution ; alternative explanations of this phenomenon , like erroneous de - dispersion and the presence of non - orthogonal polarization modes , are excluded .
if the observed , phase - resolved rm variations are common amongst pulsars , then many of the previously measured pulsar rms may be in error by as much as a few tens of rad m@xmath1 . |
by the recent global analyses with the experimental data for polarized deep inelastic scattering ( dis ) , the polarized quark and antiquark distributions are determined well @xcite .
these distributions are obtained with enough accuracy to indicate that the quark spin content is smaller than prediction of naive quark model ; @xmath4 . these polarized parton distribution functions ( pdfs )
reproduce well the experimental data ; however , the polarized gluon distribution @xmath5 can not be constrained because of indirect and small contribution through @xmath6 evolution and higher order correction at next - to - leading order ( nlo ) .
furthermore , pdf uncertainty estimation indicated large uncertainty of the gluon distribution .
it implies difficulty of the @xmath5 determination with only the polarized dis data . as a probe for the polarized gluon distribution , prompt photons
will be measured by the longitudinally polarized proton - proton collider at rhic @xcite .
the gluon distribution contributes directly in the quark - gluon compton process @xmath7 at leading order ( lo ) , and the process dominates in the whole @xmath8 region .
the future asymmetry data contain useful information for clarifying the gluon contribution to the nucleon spin .
therefore , we are interested in the influence of the prompt photon data on the @xmath5 determination by the polarized pdf analysis . in this paper , we consider constraint of prompt photon data on the polarized gluon distribution . for evaluating the data constraint
, we will compare the uncertainty of the spin asymmetry with the expected statistical error by the rhic experiments . the asymmetry uncertainty coming from the polarized pdfs
is estimated by the hessian method , and it is comparable with the measurement error . in this comparison ,
the statistical error of the spin asymmetry plays a role of constriction for the @xmath9 uncertainty via the asymmetry . in practice
, it is indicated that uncertainties of the polarized pdfs can be reduced by including new precise data for the polarized dis in the asymmetry analysis collaboration ( aac ) @xcite .
therefore , the same thing is expected by including the future data for prompt photon production .
the spin asymmetry @xmath10 is defined as a ratio of the polarized and unpolarized cross sections : @xmath11 . by the factorization theorem , the polarized cross section @xmath12 as a function of the transverse momentum @xmath8 is expressed by @xmath13 where @xmath14 is the polarized pdf of the parton @xmath15 .
we choose the aac03 pdf set @xcite .
@xmath16 is the partonic cross section ( @xmath17 ) . in order to reduce theoretical uncertainty from the scale dependence of the cross section ,
the nlo corrections are taken into account .
the nlo partonic cross sections for prompt photon production are completely known @xcite .
the renormalization and factorization scales are chosen the scale @xmath18 .
in addition , the cross section is integrated over the rapidity bin @xmath19 , which corresponds to the acceptance of the phenix detector . the unpolarized cross section @xmath20 is similarly computed with unpolarized pdfs and partonic cross sections @xcite .
we choose the grv98 unpolarized pdf set @xcite , which is also used in the aac03 analysis .
these cross sections are numerically calculated at center - of - mass ( c.m . )
energy @xmath3 and 500 gev , respectively . in this study ,
the contribution from fragmentation is neglected .
the contribution is associated with the collinear process for a scattered parton into a photon , and it can be diminished by using an isolation cut on the measured photon @xcite .
and so we should consider the isolation cut in this analysis @xcite . for examining an effect on the polarized pdf uncertainties
, we calculate the cross section for inclusive direct photon production process .
the asymmetry uncertainty is obtained from uncertainty of the polarized cross section : @xmath21 .
the uncertainty @xmath22 comes from the polarized pdf uncertainties , and can be estimated by the hessian method .
the method based on eigenvectors of the diagonalized hessian matrix is developed by cteq collaboration @xcite , and it is applied to estimate uncertainties of unpolarized pdfs @xcite .
we used the basic method , which is generally used by the polarized pdf analyses @xcite .
the uncertainty is given by @xmath23 ^ 2= \delta \chi^2 \sum_{i , j } \left ( \frac{\partial \delta \sigma^\gamma(p_{_t } ) } { \partial a_i } \right ) h_{ij}^{-1 } \left ( \frac{\partial \delta \sigma^\gamma(p_{_t } ) } { \partial a_j } \right ) \
, \label{eq : erroe - m}\ ] ] where @xmath24 are optimized parameters in the polarized pdfs .
@xmath25 is the hessian matrix which has information on the parameter errors and correlation between each parameter .
the gradient terms of the cross section @xmath26 are obtained as follows : @xmath27 \\
\nonumber & \times \frac{d \delta \hat{\sigma}_{ab}^\gamma}{d p_{_t } d\eta } ( x_a , x_b,\sqrt{s},p_{_t},\eta,\mu_{_r},\mu_{_f } ) \ .\end{aligned}\ ] ] the gradient terms of the polarized pdfs can be derived analytically at initial scale , and these terms are numerically evolved to arbitrary scale @xmath28 by the dglap equation .
furthermore , the value of @xmath29 determines a confidence level of the uncertainty .
it is obtained so that the level corresponds to 1@xmath30 error of the normal distribution @xcite .
the uncertainty therefore can be directly compared with the statistical error of experimental data . for comparison with the asymmetry uncertainty ,
the statistical error of the spin asymmetry is estimated by @xmath31 where @xmath32 is the beam polarization , @xmath33 is the integrated luminosity , and @xmath30 is the unpolarized cross section integrated over the @xmath8 bin . in this study , the @xmath30
is computed by the bin size of 5 gev interval , and it is taken the same bin size for both c.m .
energies @xmath3 and @xmath34 gev .
furthermore , other values are chosed design values at rhic @xcite : @xmath35 and @xmath36 pb@xmath37 for @xmath38 gev .
comparison of the predicted spin asymmetries by different polarized pdf sets ; aac03(nlo ) , bb ( iset=3 ) , grsv01 ( standard scenario ) , and lss ( @xmath39 scheme ) .
the dotted curves show the asymmetry uncertainty from the pdf uncertainties of the aac03 set . ]
first , we discuss predicted spin asymmetry and its uncertainty at @xmath3 gev . in fig .
[ fig : asymmetry ] , the spin asymmetry by the aac03 pdf set is compared to those by polarized pdf sets of bb ( iset=3 ) @xcite , grsv01 ( standard scenario)@xcite , and lss ( @xmath39 scheme ) @xcite .
these analyses used almost the same experimental data sets for the polarized dis , and they obtained good agreements with the data .
however , there are significant differences of the gluon distributions among them .
these differences are obviously reflected in variations of the predicted asymmetries .
the prompt photon process is sensitive to the behavior of the gluon distribution .
moreover , the asymmetry uncertainty is indicated in the same figure .
dotted curves show the uncertainty which comes from the polarized pdf uncertainties obtained by the aac analysis with the polarized dis data .
we find that these predicted asymmetries are within the large uncertainty .
these variations are caused by weak constraint of the polarized dis data on the gluon distribution .
the prompt photon data therefore are required for reducing this asymmetry uncertainty . in order to evaluate the gluon contribution to the asymmetry uncertainty , we compute the asymmetry uncertainty excluding the @xmath5 uncertainty by assuming @xmath40 in eq .
( [ eq : gradf ] ) .
figure [ fig : dgoff ] shows the asymmetry uncertainties for @xmath41 gev .
the solid curves show the current uncertainty by the aac analysis , and the dotted curves do the asymmetry uncertainty except the @xmath5 uncertainty .
the significant reduction of the uncertainty indicates that the @xmath5 uncertainty is the dominant contribution to the current uncertainty .
in addition , the asymmetry uncertainty is compared to the expected statistical errors at rhic .
the current uncertainty is much larger than these statistical errors .
if these data are included in the global analysis , the uncertainty is roughly reduced to these errors .
this suggests that the @xmath5 uncertainty is mainly improved .
the prompt photon data therefore have strong constraint on the gluon distribution .
comparison of the asymmetry uncertainty @xmath42 with the expected statistical errors for @xmath3 gev .
the dashed curves show the asymmetry uncertainty except the gluon uncertainty : @xmath43 . ] on the other hand , the data constraint on the quark and antiquark distributions is very weak .
this is because that the asymmetry uncertainty without the @xmath5 uncertainty , which is composed of the @xmath44 and @xmath45 uncertainties , is significantly less than the statistical errors .
therefore , the data with such errors do not directly affect improvements of these uncertainties .
however , as far as the antiquark is concerned , the uncertainty can be indirectly reduced because the antiquark distribution is strongly correlated with the gluon distribution in the global analysis . in practice ,
reduction of the antiquark uncertainty via the error correlation is indicated by the analysis with the fixed @xmath46 at initial scale @xcite .
this fact suggests that the constraint on the gluon distribution indirectly affects the @xmath45 determination .
in particular , it is not neglected when we perform flavor decomposition of the antiquark distributions .
next , we estimate a constraint factor for the @xmath5 uncertainty . by multiplying the gradient terms for the gluon @xmath47 by the factor in eq .
( [ eq : gradf ] ) , the constricted uncertainty of the asymmetry is defined . from comparison of the asymmetry uncertainty with the expected statistical errors for @xmath3 in the region @xmath48 gev ,
the obtained factor is @xmath49 . in this study ,
the factors for the @xmath44 and @xmath45 uncertainties are neglected .
this is simply because that these uncertainties are not directly constricted by these data .
moreover , the correlation effect on the @xmath45 uncertainty is not taken into account .
the effect can not be evaluated without including experimental data in the global analysis . since the @xmath45 contribution to the asymmetry uncertainty is already small
, the uncertainty will be slightly modified in this process . in fig .
[ fig : constrict ] , the constricted asymmetry uncertainties are compared with the expected statistical errors for @xmath3 and @xmath34 gev , respectively .
the solid curves show the asymmetry uncertainties which are obtained from the @xmath5 uncertainty multiplied by the constraint factor , and involve the @xmath44 and @xmath45 uncertainties .
if @xmath50 can be approximated by the bjorken @xmath51 around central rapidity region , the data for @xmath3 gev in the region
@xmath48 gev constrain the gluon distribution in the range @xmath52 .
although this comparison is in ideal condition that these data are put on the predicted asymmetry , this fact agrees with the results of the trial analysis including pseudo - data for @xmath0 in ref .
these data are useful in determining the gluon distribution . in the region
@xmath53 gev , these data have rather weak constraint since the unpolarized cross section rapidly decreases as @xmath8 increases .
the statistical errors depend on the @xmath8 bin size computing @xmath30 in eq .
( [ eq : staerr ] )
. we should be careful about taking the bin size for the high @xmath8 data in order to constrain equally the gluon distribution over a wide @xmath54 region .
comparison of the constricted asymmetry uncertainties with the expected statistical errors for @xmath3 and @xmath34 gev , respectively .
the solid curves show the constricted asymmetry uncertainty with the factor @xmath49 for the @xmath5 uncertainty . ] for the comparison at @xmath55 gev , we find similar behavior . in the region @xmath48 gev , the asymmetry uncertainty roughly corresponds to the statistical errors .
this indicates that these data have the same constraint as those for @xmath3 gev , and constrain the gluon distribution in the range @xmath56 . above the region ,
the statistical errors are larger than the asymmetry uncertainty .
it is noteworthy to mention here that the data constraint for @xmath55 gev is weaker than that for @xmath3 gev in the region @xmath48 gev in spite of covering the same @xmath57 region . the unpolarized cross sections for @xmath3 and 500 gev .
] the reason for the weak constraint is that the unpolarized cross section for @xmath55 gev is less than that for @xmath3 gev in the same @xmath57 region , and the integrated luminosity is still insufficient to provide the enough constraint at high @xmath8 .
figure [ fig : xsec_xt ] shows the comparison of unpolarized cross sections for both c.m .
these cross sections are indicated as a function of @xmath57 . in the region
@xmath58 , the cross section for @xmath55 gev is below one for @xmath59 gev , and indicates similar behavior . in order to obtain equal constraint at the same @xmath57
, we need more luminosity than the design value at @xmath55 gev .
therefore , the experimental data for @xmath55 gev are required as constraint on the gluon distribution at low @xmath54 .
the medium-@xmath54 behavior should be determined by using the data for @xmath59 gev . the polarized gluon distribution and its uncertainty at @xmath60 gev .
shaded area is the constricted gluon uncertainty . ] finally , let us turn to the @xmath5 uncertainty from the prompt photon data at rhic .
figure [ fig : xdg ] shows the polarized gluon distribution and its uncertainties at @xmath60 gev .
the solid curves show the @xmath5 uncertainty from the polarized dis data , and the shaded area shows the constricted uncertainty by comparison with the expected statistical errors .
the uncertainty estimated by the constraint factor is reliable in the range @xmath61 . as a practical estimation
, the uncertainty broadens gradually in the low- and medium-@xmath54 regions where data do not exist .
moreover , we note that increasing the asymmetry uncertainty with @xmath8 is due to uncertainty of the ratio of the polarized and unpolarized gluon distributions : @xmath62 .
the uncertainty significantly increases with @xmath54 due to lack of precise data for polarized dis at large @xmath54 .
the @xmath5 uncertainty becomes the same order of magnitude as the quark uncertainty from the dis data . by including future asymmetry data for the prompt photon process
, the gluon distribution can be obtained with sufficient accuracy . in this study ,
polarization of the gluon distribution is not discussed . the current uncertainty in fig .
[ fig : xdg ] indicates that we can not rule out the possibility of zero or negative polarization . as another probe for the gluon distribution , the double spin asymmetry for @xmath63 production
has recently been reported by the phenix collaboration @xcite .
since @xmath64 subprocess dominates at low @xmath8 , the cross section depends on @xmath65 ; therefore , the process is not sensitive to the sign of the gluon polarization . on the other hand ,
the prompt photon production which the qg compton process dominates is sensitive to the sign in the whole @xmath8 region . the gluon polarization is obviously reflected in the spin asymmetry . in this sense , the role of prompt photon data is of prime importance for determination of the gluon polarization .
in this paper , we have considered the uncertainty of the polarized gluon distribution for prompt photon production at rhic .
the uncertainty of the double spin asymmetry is estimated by the hessian method .
the asymmetry uncertainty mostly comes from the @xmath5 uncertainty .
the large uncertainty implies the weak constraint of the polarized dis data on the gluon distribution . by comparison with the expected statistical errors at rhic ,
we indicate that the prompt photon data have the strong constraint on it .
furthermore , we suggest that the prompt photon data in the region @xmath48 gev effectively constrain the gluon distribution . the data of both c.m .
energies constrain it in the following regions : @xmath56 at @xmath55 gev , and @xmath52 at @xmath59 gev . for clarifying the gluon contribution @xmath66 ,
the data covering a wide range of @xmath54 are required .
these experiments therefore play an important role in the @xmath5 determination .
the author would like to thank n. saito for helpful discussion and useful comments . 00 m. hirai , s. kumano , and n. saito ( aac ) , ; http://spin.riken.bnl.gov / aac/. j. blmlein and h. bttcher , . e. leader , a. v. sidorov , and d. b. stamenov , . m. glck , e. reya , m. stratmann , and w. vogelsang , .
g. bunce , n. saito , j. soffer , and w. vogelsang , ann .
nucl . part .
50 ( 2000 ) 525 . l. e. gordon and w. vogelsang , .
m. glck , e. reya , and a. vogt , .
s. frixione , .
s. frixione and w. vogelsang , .
cteq collaboration , j. pumplin , .
cteq collaboration , j. pumplin , jhep 0207 ( 2002 ) 012 ; a. d. martin , r. g. roberts , w. j. stirling , and r. s. thorne , . m. stratmann and w. vogelsang , .
phenix collaboration , s.s adler , _ et al .
_ , hep - ex/0404027 . | constraint of prompt photon data on the polarized gluon distribution is discussed in terms of uncertainty estimation for polarized parton distribution functions ( pdfs ) . by comparing uncertainty of the double spin asymmetry
@xmath0 with expected statistical errors at rhic , we found that the gluon distribution is effectively constrained in the region @xmath1 with the data at transverse momentum @xmath2 gev for center - of - mass energies @xmath3 and 500 gev .
polarized parton distribution , uncertainty estimation , prompt photon 13.60.hb , 13.85.qk , 13.88.+e |
in recent years the study of the higher dimensional einstein equations has led to interesting new types of solutions .
study of the generalized weyl solutions by emparan and reall @xcite , for instance , has led to the construction of black ring solutions @xcite and their generalizations ( see e.g. @xcite for a review ) .
the generalized weyl solutions have been classified in terms of their rod structure @xcite .
they also include multi black hole solutions , analyzed in more detail by tan and teo @xcite , as well as sequences of kaluza - klein ( kk ) bubbles and black holes , analyzed by elvang and horowitz @xcite and elvang , harmark and obers @xcite .
the presence of kk bubbles leads to interesting features for the black hole solutions .
black hole solutions with spherical horizon topology can , for instance , become arbitrarily large in the presence of a bubble @xcite .
in contrast , because of the finite size of the circle at infinity , without a bubble such kk black holes are limited in size , as demonstrated by kudoh and wiseman @xcite .
the bubbles also provide the means , to hold the black holes apart , allowing for multi black hole spacetimes without conical singularites .
small pieces of bubble are sufficient to hold large black holes in equilibrium , i.e. , they compensate the gravitational attraction of the black holes and at the same time prevent them from merging @xcite . recently , kastor , ray and traschen @xcite also addressed the role of kk bubbles for the laws of black hole mechanics . introducing the notion of bubble surface gravity , they obtained interesting relations , displaying an interchange symmetry between bubble and black hole properties .
here we address the question as to how the properties of such kaluza - klein black hole - bubble spacetimes are affected by the presence of charge .
in particular , we consider the 5 dimensional einstein - maxwell - dilaton ( emd ) gravity with arbitrary dilaton coupling parameter @xmath0 . in the limit @xmath1 , einstein - maxwell theory is recovered .
we here focus on the solution of two charged black holes on a kaluza - klein bubble . in section 2 recall the action and equations of motion .
we present the exact solution of two charged black holes on a kaluza - klein bubble in section 3 .
we analyze the horizon and global properties of this solution in section 4 . in section 5
we derive smarr - like formulae for the mass and tension and present the first law .
we address the extremal limit in section 6 and compare with charged black strings in section 7 .
we present our conclusions in section 8 .
we consider the 5-dimensional einstein - maxwell - dilaton ( emd ) gravity with action @xmath2 and arbitrary dilaton coupling parameter @xmath0 .
the field equations are given by @xmath3 the 5d emd gravity has a dual description in terms of the 3-form field strength @xmath4 defined by @xmath5 where @xmath6 denotes the hodge dual . in terms of the 3-form field strength @xmath4 the action
can be written as follows @xmath7 the corresponding field equations of the dual theory are @xmath8
as an explicit example of the sequences of kaluza - klein ( kk ) bubbles and black holes @xcite , let us consider the solution describing two vacuum black holes on a kaluza - klein bubble discussed by elvang and horowitz @xcite .
this vacuum solution has metric @xmath9 where @xmath10 and @xmath11 the metric functions @xmath12 , @xmath13 and @xmath14 may be regarded as newtonian potentials produced by line masses along the @xmath15-axis with densities @xmath16 .
these effective sources or rods characterize the solution @xcite .
the rod structure of this solution is shown in fig . 1 .
the two finite rods @xmath17 and @xmath18 for the @xmath19 coordinate correspond to event horizons of the spacetime , while the finite rod @xmath20 for the @xmath21 coordinate corresponds to the bubble .
the two semi - infinite rods @xmath22 and @xmath23 for the @xmath24 coordinate then ensure , that all sources add up to an infinite rod along the @xmath15-axis , as required @xcite . using the method of @xcite
we generate a new emd solution from the above vacuum solution @xcite .
this solution then describes two charged @xmath25-black holes sitting on a kaluza - klein bubble in the emd gravity . for @xmath26 it reduces to a solution of einstein - maxwell gravity .
the spacetime metric of the solution is given by the line element @xmath27^{2\over 1+\alpha^2_{5 } } } dt^2 \nonumber \\ & & + [ \cosh^2\gamma -e^{2u_{t}}\sinh^2\gamma]^{1\over 1+\alpha^2_{5 } } \left[e^{2u_{\psi}}d\psi^2 + e^{2u_{\phi}}d\phi^2 + e^{2\nu}(d\rho^2 + dz^2 ) \right]\end{aligned}\ ] ] the electric potential @xmath28 and the dilaton field are given by @xmath29^{\alpha_{5}^2\over 1 + \alpha^2_{5 } } \end{aligned}\ ] ] where @xmath30 and @xmath31 is an arbitrary parameter .
the maxwell 2-form is respectively @xmath32
in order to study the asymptotic behaviour of the solution we introduce the asymptotic coordinate @xmath33 defined by @xmath34 then in the asymptotic limit @xmath35 we find @xmath36,\\ & & g_{\psi\psi}\approx \rho^2 \left(1 - { 1\over r}\left[(a+c-2b){\sinh^2\gamma\over 1 + \alpha^2_{5 } } -(a+c)\right ] \right ) , \\ & & \phi \approx { \sqrt{3}\over 2}{\cosh\gamma\sinh\gamma\over \sqrt{1 + \alpha^2_{5 } } } { ( a+c-2b)\over r } , \\ & & e^{-\alpha\varphi } \approx 1 + { \alpha^2_{5 } \sinh^2\gamma\over 1 + \alpha^2_{5 } } { ( a+c-2b)\over r}.\end{aligned}\ ] ] the leading asymptotic of the metric is @xmath37 in other words , asymptotically the space - time is @xmath38 .
the compact dimension is parameterized by @xmath21 . in what follows we will consider spacetimes with a fixed length @xmath39 of the kauza - klein circle at infinity .
the rod structure of the solution in general entails conical singularities . to cure such conical singularities at the location of the rods ,
the associated coordinate must be periodic with a particular period .
* for the semi - infinite rods @xmath40 $ ] and @xmath41 corresponding to the @xmath42 part of the metric regularity requires the period @xmath43 * for the finite rod @xmath44 $ ] corresponding to the @xmath45 part of the metric regularity requires the period @xmath46 where @xmath39 is the length of the kk circle at infinity .
the first black hole horizon is located at @xmath47 for @xmath48 .
the metric of the spatial cross section of the horizon is given by @xmath49 .
\nonumber\end{aligned}\ ] ] the orbits of @xmath21 and @xmath24 shrink to zero size at @xmath50 and @xmath51 , respectively .
this shows that the black hole horizon topology is @xmath25 . by the direct calculation we find the horizon area @xmath52 the temperature of the horizon
can be obtained from the surface gravity and is given by @xmath53 the near horizon geometry of the left black hole can be illustrated by performing the coordinate transformation @xcite @xmath54 the near horizon metric then becomes @xmath55 \right .
\nonumber \\ & & \hspace{3.5 cm } \left .
+ { f_a(\theta)\over 2 f_b^2(\theta ) } r^2 \cos^2 \theta d \phi^2 + { 2 f_a^2(\theta)\over f_b(\theta ) } r^2 \sin^2 \theta d \psi^2 \right\}\end{aligned}\ ] ] with @xmath56 corresponding to a distorted black hole .
the second black hole horizon is located at @xmath47 for @xmath57 .
the metric of its spatial cross section is @xmath58.\end{aligned}\ ] ] the orbits of @xmath21 and @xmath24 shrink to zero size at @xmath59 and @xmath60 , respectively showing that the horizon topology is @xmath25 .
the horizon area is @xmath61 and its temperature @xmath62 the near horizon metric for the second black hole can be found from the near horizon metric of the first black hole by interchanging the parameters @xmath63 and @xmath64 .
the near horizon metric shows that the second black holes is distorted by the interaction with the first black holes and the kaluza - klein bubble .
the bubble is located at @xmath47 for @xmath65 .
the metric on the bubble for @xmath66 and @xmath47 is given by @xmath67^{1\over 1 + \alpha^2_5 } \left[4(a - z)(z+c)d\psi^2 \right.\nonumber \\ \\ \left.+ { 4b^2 ( a+c)^2 \over ( a+b)(b+c ) } { dz^2\over b^2-z^2 } \right ] .
\nonumber\end{aligned}\ ] ] as it is clear form the rod structure diagram and as one sees from the bubble metric , the orbits of @xmath24 do not vanish at @xmath68 .
so the bubble has cylindrical topology .
the proper distance along a curve with @xmath69 between the black holes is given by @xmath70^{1\over 2(1 + \alpha^2_5 ) } { dz\over \sqrt{b^2-z^2 } } .\end{aligned}\ ] ] it can be shown that @xmath71 in order to compute the adm mass of the solution we use the general results of @xcite and @xcite .
we then find @xmath72 \nonumber \\ & = & 4\pi { b(a+c)\over \sqrt{(a+b)(b+c ) } } \left[a+c - b + { 3\sinh^2\gamma\over 2(1+\alpha^2_{5})}(a+c-2b ) \right ] \label{mass}\end{aligned}\ ] ] with the asymptotic form of the relevant metric components given by @xmath73 the mass ( [ mass ] ) can be directly read off from the asymptotics of the metric potentials @xcite @xmath74 the adm mass can also be calculated via the generalized komar integral introduced in @xcite @xmath75 where @xmath76 is the surface element and @xmath77 , @xmath78 are the time translation and the killing vector associated with the fifth dimension , respectively . here
the integration is performed over a 2-dimensional sphere at spatial infinity of @xmath79 . when dealing with multi - object configurations it is useful to properly define the intrinsic mass of each object in the configuration .
as in four dimensions , the natural definition of the intrinsic mass of each black hole is the komar integral evaluated on the horizon of the corresponding black hole @xmath80 here the integration is over a 2-dimensional surface which is an intersection of the horizon with a constant @xmath19 and @xmath21 hypersurface .
the intrinsic masses of the black holes for the solution under consideration are given by @xmath81 @xmath82 where @xmath83 is the surface gravity of the @xmath84-th black hole . in the same way we can define the intrinsic mass of the kaluza - klein bubble
@xmath85 the integration is over the 2-dimensional boundary of a constant @xmath19 and @xmath21 hypersurface at the bubble .
it is worth noting that the intrinsic mass of the bubble is given by @xmath86 where @xmath87 is the so - called bubble surface gravity and @xmath88 is the bubble area .
these quantities have been introduced in @xcite and will be discussed in more detail in the next section . in our case
the bubble is smooth , i.e. , there are no conical singularities at the bubble . for such bubbles one obtains @xcite @xmath89 and @xmath90 where @xmath39 is the length of the kaluza - klein circle at infinity .
note , that @xmath87 and @xmath88 are not affected by the presence of charge .
the electric charge of the @xmath84-th black hole is defined by @xmath91 explicit calculations give @xmath92 thus the individual black holes have the same charge to mass ratio . denoting by @xmath93 the electric potential evaluated on the horizon @xmath94
, we obtain explicitly @xmath95 finally , we define the dilaton charge @xmath96 via the asymptotic behaviour of the dilaton field @xmath97 comparison with the electric charges and potential then yields the relation @xmath98 the dilaton charge thus represents no independent parameter .
it can be checked that our solution satisfies the following smarr - like relation @xmath99 here eq .
( [ sigma ] ) can be used to replace the electromagnetic contributions by the dilaton contribution .
below we give a general proof of the smarr - like relation ( [ smarr ] ) .
our starting point is the generalized komar integral ( [ komar ] ) . using the gauss theorem we find that the adm - mass can be written as a sum of a bulk integral over a constant @xmath19 and @xmath21 hypersurface and surface integrals on the black holes horizons and the bubble @xmath100 -\sum_{i}{l\over 16\pi } \oint_{{\cal h}_i } ds^{\mu\nu\sigma}\left(2\eta_{\sigma}\nabla_{\mu}\xi_{\nu } + \xi_{\nu}\nabla_{\mu}\eta_{\sigma}\right ) \nonumber \\ \\ & & -{l\over 16\pi } \oint_{bubble } ds^{\mu\nu\sigma}\left(2\eta_{\sigma}\nabla_{\mu}\xi_{\nu } + \xi_{\nu}\nabla_{\mu}\eta_{\sigma}\right ) .
\nonumber\end{aligned}\ ] ] on the horizons we have @xmath101 } d\omega,\end{aligned}\ ] ] where @xmath102 is the magnitude of the 2-dimensional surface element and @xmath103 is the second null normal to the horizon normalized such as @xmath104 . substituting into the horizon surface integrals and taking into account that @xmath105 and @xmath106 are commuting we obtain @xmath107 and therefore @xmath108 by definition , on the horizons @xmath109 , where @xmath83 is the surface gravity of the @xmath84-th horizon .
hence we obtain @xmath110 and therefore @xmath111 where we have made use of the fact that the surface gravity is constant on the horizon .
further , taking into account that @xmath112 is the area of the @xmath84-th horizon we find that @xmath113 let us denote by @xmath114 and @xmath115 the unit spacelike normals to the bubble surface .
the bubble surface gravity is defined by the following equality on the bubble ( see @xcite ) @xmath116}=\kappa_{b}(n^{1}_{\mu}n^2_{\nu}-n^{1}_{\nu}n^2_{\mu } ) .\end{aligned}\ ] ] as shown in @xcite , the bubble surface gravity @xmath87 is constant on the bubble surface . in our case
this is obvious since @xmath117 . for the bubble surface element we can write @xmath118}d\omega.\end{aligned}\ ] ] substituting in the bubble surface
integral and taking into account that @xmath106 vanishes on the bubble we have @xmath119 and therefore we find @xmath120 where @xmath121 is the bubble area as defined in @xcite .
let us now evaluate the bulk integral @xmath122.\end{aligned}\ ] ] using the field equations ( [ emdfe ] ) we obtain @xmath123.\end{aligned}\ ] ] making use of the lie symmetries @xmath124 and @xmath125 as well as the field equations one can show that the following relations are satisfied @xmath126\lambda}a_{\lambda } \right ) .
\nonumber\end{aligned}\ ] ] using these relations and the fact that the integrand has vanishing lie derivative with respect to the killing vector @xmath106 the bulk integral can be presented as a sum of surface integrals by applying the gauss theorem @xmath127 \nonumber \\
\nonumber\\ + { l\over 4\pi } \oint_{bubble } ds_{\mu\nu\sigma}\left[e^{-2\alpha\varphi}\left(f^{\mu\sigma}\xi^{\lambda}a_{\lambda } + \xi^{\mu}f^{\sigma\lambda}a_{\lambda } \right)\eta^{\nu } - { 1\over 2 } e^{-2\alpha\varphi}f^{\mu\sigma}\eta^{\lambda}a_{\lambda}\xi^{\nu } \right ] \\\nonumber \\ - { l\over 4\pi } \oint_{\infty } ds_{\mu\nu\sigma}\left[e^{-2\alpha\varphi}\left(f^{\mu\sigma}\xi^{\lambda}a_{\lambda } + \xi^{\mu}f^{\sigma\lambda}a_{\lambda } \right)\eta^{\nu } - { 1\over 2 } e^{-2\alpha\varphi}f^{\mu\sigma}\eta^{\lambda}a_{\lambda}\xi^{\nu } \right ] . \nonumber\end{aligned}\ ] ] the killing vector @xmath106 vanishes on the bubble and therefore the surface integral on the bubble vanishes too .
the same holds for the surface integral at infinity since we consider a gauge in which @xmath128 vanishes at infinity . as a consequence
we find that only the first term in ( [ surfint ] ) survives .
further we consider pure electric solutions satisfying @xmath129}=0 , \ ; \ ; \eta^{\mu}f_{\mu\nu}=0.\end{aligned}\ ] ] under these conditions and using the 2-dimensional horizon surface element ( [ hse ] ) we find that @xmath130 where we have taken into account that @xmath131 is the 3-dimensional surface element on the horizon . in the final step we use the fact that the potential @xmath132 is constant is constant on the horizon can be shown by the well - known manner .
on the horizon we have @xmath133 which , in view of the field equations , implies that @xmath134 is null on the horizon .
since @xmath135 is null on the horizon and orthogonal to @xmath134 it follows that on the horizon @xmath134 is proportional to @xmath135 , @xmath136 . for an arbitrary vector @xmath137
tangential to the horizon we have @xmath138 ( on the horizon ) which shows that @xmath28 is constant on the horizon .
] on the horizon and therefore we obtain @xmath139 in this way we have proven the smarr - like formula @xmath140 obviously the above relation can be generalized easily for the case when more than one bubble is present in the configuration .
a smarr - like relation can also be found for the tension .
the tension can be calculated by the following komar integral @xcite @xmath141 with the asymptotic form of the relevant metric components given by eq .
( [ asym ] ) , the tension can be directly read off from the asymptotics of the metric potentials @xcite @xmath142 following the same method as for the derivation of the smarr - like relation for adm mass we find @xmath143 combining this relation with the relation for adm mass one can easily show that the following relation holds @xmath144 in the absence of bubbles the first law can be derived following the well known method . in the presence of bubbles we have to find the contribution of the bubbles to the first law .
it was however shown in @xcite that the smooth bubbles ( i.e. without conical singularities ) do not give any contribution to the first law . in this paper
we consider namely smooth bubbles .
therefore , in our case the first law conserves its usual form @xmath145 let us note again that we keep @xmath39 fixed .
the extremal solutions are obtained in the limit @xmath146 , keeping the charges @xmath147 and @xmath148 fixed .
the explicit form of the extremal solutions is the following @xmath149 , \\
% \nonumber \\ \phi & = & { \sqrt{3}\over 2 \sqrt{1+\alpha^2_5 } } { { q_2\over r_2 } + { q_1\over r_3}\over 1 + { q_2\over r_2 } + { q_1\over r_3 } } , \nonumber \\\nonumber \\ e^{-\alpha\varphi } & = & \left(1 + { q_2 \over r_2 } + { q_1\over r_3}\right)^{\alpha^2_5\over 1 + \alpha^2_5 } \nonumber\end{aligned}\ ] ] where the parameters @xmath150 and @xmath151 are given by @xmath152 strictly speaking , the extremal solutions are singular and they can not be interpreted as extremal black holes . as one can see the dilaton field is divergent on the candidate - horizons @xmath153 and @xmath154 for @xmath155 . as we shall show below , only in the einstein - maxwell case ( @xmath26 )
the solutions describe two extremal black holes sitting on a kaluza klein bubble .
the extremal solutions can be casted in new coordinates @xmath33 and @xmath156 by performing the transformation @xmath157 which gives @xmath158^{-{2\over 1 + \alpha^2_5 } } dt^2 + \nonumber
\\ \nonumber\\ & & \hspace{-1 cm } \left [ 1 + { q_{2}\over r - b - b\cos\vartheta } + { q_{1}\over r - b+b\cos\vartheta}\right]^{{1\over 1 + \alpha^2_5 } } \left[{dr^2\over 1-{2b\over r } } + r^2d\vartheta^2 + r^2\sin^2\vartheta d\psi^2 + ( 1 - { 2b\over r})d\phi^2 \right ] , \nonumber \\ \nonumber\\ \phi & = & { \sqrt{3}\over 2 \sqrt{1+\alpha^2_5}}\left [ 1 - { 1\over 1 + { q_{2}\over r - b - b\cos\vartheta } + { q_{1}\over r - b+b\cos\vartheta}}\right ] , \nonumber \\ \\ e^{-\alpha\varphi}&=&\left[1 + { q_{2}\over r - b - b\cos\vartheta } + { q_{1}\over r - b+b\cos\vartheta } \right]^{\alpha^2_5\over 1 + \alpha^2_5}. \nonumber\end{aligned}\ ] ] the new coordinates @xmath33 and @xmath156 appear to be more suitable from a physical point of view and the extremal solutions look more tractable . in the particular case
@xmath159 we obtain the 5d solution describing a static kk bubble in standard spherical coordinates .
the extremal solutions can be interpreted physically as solutions describing two charged objects whose gravitational attraction is balanced by the electric repulsion in the presence of a kaluza - klein bubble .
let us now show that , in the einstein - maxwell case , the extremal solutions describe two extremal black holes on a kaluza - klein bubble .
for this purpose we consider the metric in the vicinity of the candidate - horizon @xmath160 and put @xmath161 where @xmath162 and @xmath163 are small . expanding the metric around @xmath164 and @xmath165 and performing the coordinate transformation @xmath166 where @xmath167 we obtain @xmath168 here
@xmath169 is the metric on the unit three dimensional sphere , i.e. , @xmath170 where @xmath171 and @xmath172 . since @xmath173 in the extremal limit the angle @xmath174 has the canonical period @xmath175 .
it is seen from ( [ extrxpansion ] ) that the metric remains regular as @xmath176 and @xmath177 , and the same is true for the electromagnetic field .
therefore @xmath160 is a ( degenerate ) event horizon with @xmath25 topology as follows from ( [ extrxpansion ] ) .
the radius of the horizon is @xmath178 and hence the area of the horizon is @xmath179 the area is also found from ( [ secondbharea ] ) by taking the extremal limit . in the same manner
one can show that @xmath180 is a ( degenerate ) event horizon with @xmath25 topology and area @xmath181 the extremal limit of ( [ firstbherea ] ) gives the same expression for the erea of the first extremal black hole .
for the two extremal einstein - maxwell black holes sitting on a kaluza - klein bubble we have @xmath182 when @xmath155 we consider near extremal ( ne ) solutions . we here find that the area of the ne black holes tends to zero @xmath183 the temperature of the ne black holes is @xmath184 the mass of the bubble and the adm mass are , respectively , @xmath185 the proper distance @xmath186 between the black holes becomes arbitrary large when extremality is approached .
for the vacuum solution the question was addressed , whether the total area of the two black holes @xmath187 is larger or smaller than the area of a black string @xmath188 with the same mass @xcite .
it turned out , that @xmath189 showing that the black string is entropically favoured . here
, we address this question for charged black holes . in order to generate the solution describing a 5-dimensional black string in einstein - maxwell - dilaton gravity we consider the 4-dimensional schwarzschild solution trivially embedded in five dimensions @xmath190 using the method of hep - th/0507097 and ( [ emebeddedsch ] ) as a seed solution one obtains the following emd solution @xmath191^{2\over 1 + \alpha^2_5}}dt^2 + \left[1 + { 2m\over r}\sinh^2\gamma_1\right]^{1\over 1 + \alpha^2_5}\left [ { dr^2 \over 1-{2m\over r } } + r^2d\omega^2 + dx^2_{5}\right ] , \nonumber \\ \nonumber \\ & & e^{-2\alpha\varphi } = \left[1 + { 2m\over r}\sinh^2\gamma_1\right]^{2\alpha^2_5\over 1 + \alpha^2_5 } , \\ \nonumber \\ & & \phi= { \sqrt{3}\over 2 } { \cosh\gamma_1 \sinh\gamma_1 \over \sqrt{1 + \alpha^2_5 } } { 2m\over r + 2m\sinh^2\gamma_1}\nonumber\end{aligned}\ ] ] an alternative derivation of this solution has been given in @xcite .
the adm mass , the charge , the horizon area and the temperature of the emd black string are given by @xmath192 , \nonumber \\\nonumber \\ & & q^{s } = { \sqrt{3}\cosh\gamma_1\sinh\gamma_1 \over \sqrt{1 + \alpha^2_5 } } ml , \nonumber\\ \\ & & { \cal a}^{s}_{h}= 16\pi \cosh^{3\over 1 + \alpha^2_5}\gamma_1\ , m^2 l , \nonumber\\ \nonumber \\ & & t^{s } = { \cosh^{-3\over 1 + \alpha^2_5}\gamma_1\ , \over 8\pi m } \nonumber\end{aligned}\ ] ] and the following smarr - like relation is satisfied @xmath193 the extremal limit of the black string solution is obtained for @xmath194 and @xmath195 keeping the charge @xmath196 finite .
the extremal solution is singular with zero horizon area for all coupling parameters @xmath197 .
the temperature of the near extremal solution tends to zero for @xmath198 , to infinity for @xmath199 and the temperature is finite for @xmath200 .
comparing now the area of two charged black holes @xmath187 with the area of a charged string of the same mass , charge and the length @xmath39 of the kaluza - klein circle , we find @xmath201 this shows that the solution is globally unstable as all solutions of this type .
we have constructed a solution of two black holes on a kaluza - klein bubble in einstein - maxwell - dilaton theory , and studied the consequences of the presence of charge for the properties of this solution .
furthermore , we have obtained a generalized smarr relation and first law for this solution , extending recent vacuum results @xcite .
while we have focussed here on a particular solution @xcite , the calculations are easily extended to general sequences of smooth kaluza - klein ( kk ) bubbles and black holes @xcite . for these
we expect the smarr relation @xmath202 and first law @xmath203 as for all solutions of the considered type the question of their classical stability remains open .
future work may involve the inclusion of rotation @xcite .
considering the solutions from the point of view of string theory , on the other hand , may necessitate the inclusion of further fields for consistency .
s.y . gratefully acknowledges support by the alexander von humboldt foundation .
r. emparan and h. s. reall , phys . rev .
d * 65 * , 084025 ( 2002 ) [ arxiv : hep - th/0110258 ] .
r. emparan and h. s. reall , phys .
* 88 * ( 2002 ) 101101 [ arxiv : hep - th/0110260 ] .
r. emparan and h. s. reall , class .
* 23 * ( 2006 ) r169 [ arxiv : hep - th/0608012 ] .
h. s. tan and e. teo , phys .
d * 68 * ( 2003 ) 044021 [ arxiv : hep - th/0306044 ] .
h. elvang and g. t. horowitz , phys .
d * 67 * ( 2003 ) 044015 [ arxiv : hep - th/0210303 ] .
h. elvang , t. harmark and n. a. obers , jhep * 0501 * ( 2005 ) 003 [ arxiv : hep - th/0407050 ] . h. kudoh and t. wiseman , phys .
* 94 * ( 2005 ) 161102 [ arxiv : hep - th/0409111 ] .
d. kastor , s. ray and j. traschen , class .
* 25 * ( 2008 ) 125004 [ arxiv:0803.2019 [ hep - th ] ]
. s. s. yazadjiev , arxiv : hep - th/0507097 .
s. deser and m. soldate , nucl .
b * 311 * ( 1989 ) 739 .
p. k. townsend and m. zamaklar , class .
* 18 * ( 2001 ) 5269 [ arxiv : hep - th/0107228 ] . t. harmark and n. a. obers , class .
* 21 * ( 2004 ) 1709 [ arxiv : hep - th/0309116 ] .
b. kol , e. sorkin and t. piran , phys .
d * 69 * ( 2004 ) 064031 [ arxiv : hep - th/0309190 ] .
t. harmark and n. a. obers , jhep * 0405 * ( 2004 ) 043 [ arxiv : hep - th/0403103 ] .
b. kleihaus , j. kunz and e. radu , jhep * 0606 * ( 2006 ) 016 [ arxiv : hep - th/0603119 ] .
s. tomizawa , h. iguchi and t. mishima , arxiv : hep - th/0702207 . | we construct a solution of two black holes on a kaluza - klein bubble in einstein - maxwell - dilaton theory .
we explore the consequences of the presence of charge for the properties of this solution , and obtain a generalized smarr relation and first law . |
consider the canonical non - parametric regression setup @xmath0 where @xmath1 is an unknown function in @xmath2 $ ] , @xmath3 and the @xmath4 s are i.i.d .
standard gaussian random variables .
we assume the noise level @xmath5 is known and , without loss of generality , set @xmath6 throughout .
based on observing @xmath7 , estimating @xmath1 or various functionals of @xmath1 has been the central problem in non - parametric function estimation .
the asymptotic optimality of estimators is usually associated with the optimal rate of convergence in terms of minimax risk . a huge body of literature has been devoted to the evaluation of minimax risks under @xmath8 loss over certain function spaces ; see , for example , pinsker @xcite , ibragimov and hasminskii @xcite , golubev and nussbaum @xcite , efroimovich @xcite , belitser and levit @xcite and goldenshluger and tsybakov @xcite .
an excellent survey of the literature in this area can be found in efromovich @xcite . sometimes , instead of estimating @xmath1 itself , one is interested in making statistical inference about future observations from the same process that generated @xmath9 .
a predictive distribution function assigns probabilities to all possible outcomes of a random variable .
it thus provides a complete description of the uncertainty associated with a prediction .
the minimaxity of predictive density estimators has been studied for finite - dimensional parametric models ; see , for example , liang and barron @xcite , george , liang and xu @xcite , aslan @xcite and george and xu @xcite .
however , so far , few results have been obtained on predictive density estimation for non - parametric models .
the major thrust of this paper is to establish the asymptotic minimax risk for predictive density estimation under kullback
leibler loss in the context of non - parametric regression .
our result closely parallels the well - known work by pinsker @xcite for non - parametric function estimation under @xmath8 loss and provides a benchmark for studying the optimality of density estimates for non - parametric regression .
let @xmath10 denote a vector of future observations from model ( [ eq : x ] ) at locations @xmath11 .
to evaluate the performance of density prediction across the whole curve , we assume that the @xmath12 s are equally spaced dense ( that is , @xmath13 ) grids in @xmath14 $ ] . given @xmath1 , the conditional density @xmath15 is a product of @xmath16 , where @xmath17 denotes a univariate gaussian density function with mean @xmath18 and unit variance .
based on observing @xmath19 , we estimate @xmath15 by a predictive density @xmath20 , a non - negative function of @xmath21 that integrates to 1 with respect to @xmath21 .
common approaches to constructing @xmath20 includes the `` plug - in '' rule that simply substitutes an estimate @xmath22 for @xmath1 in @xmath15 , @xmath23 and the bayes rule that integrates @xmath1 with respect to a prior @xmath24 to obtain @xmath25 we measure the discrepancy between @xmath15 and @xmath20 by the average kullback
leibler ( kl ) divergence @xmath26 assuming that @xmath1 belongs to a function space @xmath27 , such as a sobolev space , we are interested in the minimax risk @xmath28 it is worth observing that in this framework , the densities of future observations @xmath29 are estimated simultaneously by @xmath20 .
an alternative approach is to estimate the densities individually by @xmath30 with risk @xmath31 when the @xmath12 s are equally spaced and @xmath32 goes to infinity , the risk above converges to @xmath33 which can be interpreted as the integrated kl risk of prediction at a random location @xmath34 in @xmath14 $ ] .
this individual prediction problem can be studied in our simultaneous prediction framework with @xmath20 restricted to a product form , that is , @xmath35 .
for example , the plug - in estimator ( [ hp : plug - in ] ) has such a product form and it is easy to check that its individual estimation risk ( [ risk : individual ] ) is the same as its simultaneous estimation risk ( [ risk : simultaneous ] ) .
in general , simultaneous prediction considers a broader class of @xmath36 than the one considered by individual prediction .
therefore , simultaneous prediction is more efficient since the corresponding minimax risk ( [ minimax : f ] ) is less than or equal to the one with individual prediction .
this is distinct from estimating @xmath1 itself under @xmath8 loss where , due to the additivity of @xmath8 loss , simultaneous estimation and individual estimation are equivalent .
this paper is organized as follows . in section [ sec2 ] , we show that the problem of predictive density estimation for a non - parametric regression model can be converted to the one for a gaussian sequence model with a constrained parameter space
. direct evaluation of the minimax risk is difficult because of the constraint on the parameter space .
therefore , in section [ sec3 ] , we first derive the minimax risk over a special class of @xmath36 that consists of predictive densities under gaussian priors on the unconstrained parameter space @xmath37 .
then , in section [ sec4 ] , we show that this minimax risk is asymptotically equivalent to the overall minimax risk .
finally , in section [ sec5 ] , we provide two explicit examples of minimax risks over @xmath8 balls and sobolev spaces .
let @xmath38 be the orthonormal trigonometric basis of @xmath2 $ ] , that is , @xmath39 then , @xmath40 , where @xmath41 is the coefficient with respect to the @xmath42th basis element @xmath43 . a function space
@xmath27 corresponds to a constraint on the parameter space of @xmath44 . in this paper
, we consider function spaces whose parameter spaces @xmath45 have ellipsoid constraints , that is , @xmath46 where @xmath47 and @xmath48 .
we approximate @xmath1 by a finite summation @xmath49 .
the bias incurred by estimating @xmath50 instead of @xmath51 can be expressed as @xmath52 ^ 2 = \frac{1}{2 m } \sum_{i = n+1}^{\infty } \theta_i^2.\end{aligned}\ ] ] this bias is often negligible compared to the prediction risk ( [ risk : simultaneous ] ) ; for example , it is of order @xmath53 for sobolov ellipsoids @xmath54 , as defined in ( [ sobolev_space ] ) .
therefore , from now on , we set @xmath55 .
let @xmath56 , @xmath57 be a @xmath58 matrix whose @xmath59th entry equals @xmath60 and @xmath61 be a @xmath62 matrix whose @xmath59th entry equals @xmath63 . then
, @xmath64 and @xmath65 are two independent gaussian vectors with @xmath66 and @xmath67 , where @xmath68 denotes the @xmath69 identity matrix .
note that since the @xmath70 s and @xmath12 s are equally spaced , we have @xmath71 and @xmath72 . defining @xmath73
it is then easy to check that @xmath74 and @xmath75 are independent and that@xmath76 where @xmath77 and @xmath78 .
we refer to the model above as a _ gaussian sequence model _ since its number of parameters is increasing at the same rate as the number of data points .
consider the problem of predictive density estimation for the gaussian sequence model ( [ model : normal ] ) .
let @xmath79 denote a predictive density function of @xmath80 given @xmath81 .
the incurred kl risk is defined to be @xmath82 and the corresponding minimax risk is given by @xmath83 the following theorem states that the two minimax risks , the one associated with @xmath84 from a non - parametric regression model and the one associated with @xmath85 from a normal sequence model , are equivalent .
[ teo2.1 ] @xmath86 , where @xmath87 is defined in and @xmath88 in .
see the .
the idea of reducing a non - parametric regression model to a gaussian sequence model via an orthonormal function basis has been widely used for non - parametric function estimation .
early references include ibraginov and hasminskii @xcite , efromovich and pinsker @xcite and references therein
. for recent developments , see brown and low @xcite , nussbaum @xcite and johnstone @xcite .
our proof of theorem [ teo2.1 ] , given in the , implies that _ simultaneous _ estimation of predictive densities in these two models are equivalent .
however , this equivalence does not hold for the _ individual _ estimation approach described in section [ sec1 ] because the product form of the density estimators , that is , @xmath89 , is not retained under the transformation .
direct evaluation of the minimax risk ( [ def : r ] ) is difficult because the parameter space @xmath90 is constrained . in this section ,
we first consider a subclass of density estimators that have simple forms and investigate the minimax risk over this subclass . in next section , we then show that the minimax risk over this subclass is asymptotically equivalent to the overall minimax risk @xmath91 .
such an approach was first used in pinsker @xcite to establish a minimax risk bound for the function estimation problem .
it inspired a series of developments , including belitser and levit @xcite , tsybakov @xcite and goldenshluger and tsybakov @xcite .
recall that in the problem of estimating the mean of a gaussian sequence model under @xmath8 loss , diagonal linear estimators of the form @xmath92 play an important role .
indeed , pinsker @xcite showed that when the parameter space ( [ theta(c ) ] ) is an ellipsoid , the minimax risk among diagonal linear estimators is asymptotically minimax among all estimators .
moreover , the results in diaconis and ylvisaker @xcite imply that if such a diagonal linear estimator is bayes , then the prior @xmath24 must be a gaussian prior with a diagonal covariance matrix .
similarly , in investigating the minimax risk of predictive density estimation , we first restrict our attention to a special class of @xmath36 that are bayes rules under gaussian priors over the unconstrained parameter space @xmath37 . due to the above connection
, we call these predictive densities _ linear _ predictive densities and call the minimax risk over this class the _ linear _ minimax risk , even though ` linear ' does not have any literal meaning in our setting . under a gaussian prior @xmath93 , where @xmath94 and @xmath95 for @xmath96 , the linear predictive density @xmath97 is given by @xmath98 note that @xmath97 is not a bayes estimator for the problem described in section [ sec2 ] because the prior distribution @xmath99 is supported on @xmath37 instead of on the ellipsoidal space @xmath100 .
nonetheless , @xmath97 is a valid predictive density function .
the following lemma provides an explicit form of the average kl risk of @xmath97 .
[ lem3.1 ] the average kullback leibler risk ( [ risk : simultaneous ] ) of @xmath97 is given by @xmath101,\ ] ] where @xmath102 .
let @xmath103 denote the posterior predictive density under the uniform prior @xmath104 , namely , @xmath105 then , by @xcite , lemma 2 , the average kl risk of @xmath97 is given by @xmath106 where @xmath107 and @xmath108 denotes the marginal distribution of @xmath109 under the normal prior @xmath110 .
it is easy to check that @xmath111 and @xmath112 - \frac{1}{2 m } \sum_{i=1}^n \frac{\s2_{n+m } + \th_i^2}{\s2_{n+m } + s_i } , \label{lemma1:part2 } \\
e \log m_s(x ; \s2_n ) & = & -\frac{n}{2 m } \sum_{i=1}^n \log [ 2 \curpi ( \s2_n + s_i ) ] - \frac{1}{2 m } \sum_{i=1}^n \frac{\s2_n + \th_i^2}{\s2_n + s_i}. \label{lemma1:part3}\end{aligned}\ ] ] the lemma then follows immediately by combining equations ( [ eq : marginal representation])([lemma1:part3 ] ) .
we denote the linear minimax risk over all @xmath97 by @xmath113 , that is , @xmath114 this linear minimax risk is not directly tractable because the inside maximization is over a constrained space @xmath90 . in the following theorem
, we first show that we can switch the order of @xmath115 and @xmath116 in equation ( [ r_l : def ] ) and then evaluate @xmath117 using the lagrange multiplier method
. the following notation will be useful throughout .
let @xmath118 denote a solution of the equation @xmath119_+ = 2 c,\ ] ] where @xmath120_{+ } = \sup(x , 0)$ ] , and let @xmath121 be @xmath122_+\ ] ] for @xmath123 .
[ teo3.2 ] suppose that the parameter space @xmath90 is an ellipsoid , as defined in ( [ theta(c ) ] ) .
the linear minimax risk is then given by @xmath124 where @xmath121 is defined as in ( [ def : tth ] ) .
the linear minimax estimator @xmath125 is the bayes predictive density under a gaussian prior @xmath126 namely , @xmath127 with @xmath128 we first prove equality ( [ eq : solution ] ) .
it is easy to check that for any fixed @xmath44 , @xmath129 achieves its minimum at @xmath130 , and @xmath131 to calculate the maximum of the above quantity over @xmath132 , one needs to solve @xmath133 with the lagrangian @xmath134 simple calculation reveals that the maximum is attained at @xmath135 given by ( [ def : tth ] ) .
next , we prove equality ( [ eq : exchange ] ) , that is , that the order of inf and sup can be exchanged .
note that for any diagonal matrix @xmath136 , we have @xmath137 therefore , if there exists an @xmath136 such that @xmath138 then all of the inequalities in ( [ ineq : chain ] ) become equalities . if we let @xmath139 , then @xmath140 where the second equality holds because @xmath141 and @xmath142 is a solution to @xmath143 since @xmath144 implies that @xmath145 , we have @xmath146 which completes the proof . note that @xmath147 , so we have @xmath148 for @xmath149 , where
@xmath150 this implies that the prior distribution corresponding to the linear minimax estimator , that is , @xmath151 , puts a point mass at zero for @xmath152 for all @xmath153 .
in this section , we turn to establishing the asymptotic behavior of the minimax risk @xmath154 over all predictive density estimators . by definition ,
we extend the approach in @xcite to show that the difference between @xmath154 and @xmath156 vanishes as the number of observations @xmath157 goes to infinity .
therefore , the overall minimax risk is asymptotically equivalent to the linear minimax risk .
this also implies that the gaussian prior @xmath158 defined in ( [ pi_v ] ) is asymptotically least favorable .
the following lemma provides a lower bound for the overall minimax risk @xmath154 under some conditions .
[ lem4.1 ] let @xmath159 be a sequence such that for some @xmath160 , @xmath161^{1/2 } \le c.\ ] ] then , as @xmath162 , the minimax risk @xmath88 has the following lower bound : @xmath163 see the .
note that , as shown in the proof , for a posterior density with a gaussian prior @xmath164 , where @xmath94 , condition ( [ cond : s_i ] ) guarantees @xmath110 to have most of its mass inside @xmath45 , in the sense that @xmath165 for some @xmath166 . with the lower bound in the above lemma
, we are ready to prove the main result in this paper , which shows that the overall minimax risk @xmath88 is asymptotically equivalent to the linear minimax risk @xmath113 .
[ teo4.2 ] suppose that @xmath100 is the ellipsoid defined in ( [ theta(c ) ] ) and @xmath167 is defined in ( [ def : tth ] ) .
if @xmath168 and @xmath169 then @xmath170 by definition , @xmath171 .
so , to prove this theorem , it suffices to show that as , @xmath172 for a fixed constant @xmath173 , let @xmath174^{{1}/{2}}$ ] and let @xmath175 for @xmath176 .
it is easy to check that the sequence @xmath177 satisfies the condition ( [ cond : s_i ] ) .
therefore , by theorem [ lem4.1 ] , @xmath178\\[-8pt ] & = & r_l(\theta ) - \frac{1}{2 m } \sum_{i=1}^n \log \frac{(\s2_n + b_i^2)(\s2_{n+m } + \tilde{\theta}_i^2)}{(\s2_{n+m } + b_i^2)(\s2_n + \tilde{\theta}_i^2 ) } + \mathrm{o}(\s2_n^\alpha)\qquad \mbox{as } \s2_n \rightarrow 0.\nonumber\end{aligned}\ ] ] next , we will derive the convergence rate of @xmath113 and show that the other terms are of smaller order .
using the fact that @xmath179 for @xmath149 ( see ( [ def : n ] ) ) , we can rewrite @xmath113 as @xmath180 when @xmath168 , we have @xmath181 and @xmath182 .
therefore , by means of a taylor expansion , @xmath183 similarly , since @xmath184 for @xmath149 , the second term in ( [ ineq : lower_bound ] ) can be written as @xmath185 for every @xmath186 , we have @xmath187 ( \s2_{n+m } + \tilde{\theta}_i^2 ) } { [ ( 1 + \gamma ) \s2_{n+m } + \tilde{\theta}_i^2 ] ( \s2_n + \tilde{\theta}_i^2 ) } \biggr)\\ & = & \log \biggl ( 1 + \gamma \frac{(\s2_n - \s2_{n+m } ) \tilde{\theta}_i^2}{(\s2_n + \tilde{\theta}_i^2)(\s2_{n+m } + \tilde{\theta}_i^2 ) + \gamma \s2_n ( \s2_{n+m } + \tilde{\theta}_i^2 ) } \biggr ) \\ & \leq & \log \biggl ( 1 + \gamma \frac{(\s2_n - \s2_{n+m } ) \tilde{\theta}_i^2}{(\s2_n + \tilde{\theta}_i^2)\s2_{n+m } } \biggr).\end{aligned}\ ] ] again using a taylor expansion , as well as the condition that @xmath188 , we obtain @xmath189 finally , since @xmath190 , by choosing @xmath173 , the last term in ( [ ineq : lower_bound ] ) satisfies @xmath191 combining ( [ ineq : lower_bound])([eq : term_3 ] ) , the theorem then follows .
in this section , we apply theorems [ teo3.2 ] and [ teo4.2 ] to establish asymptotic behaviors of minimax risks over some constrained parameter spaces . in particular , we consider the asymptotics over @xmath8 balls and sobolev ellipsoids .
[ ex1 ] suppose that @xmath192 and @xmath44 is restricted in an @xmath8 ball , @xmath193 the @xmath8 ball can be considered as a variant of the ellipsoid ( [ theta(c ) ] ) with @xmath194 and @xmath195 .
although the values of the @xmath196 s here depend on @xmath157 , the proofs of the above theorems are still valid .
it is easy to see that @xmath197 defined in ( [ def : n ] ) is equal to @xmath157 and that @xmath198
. therefore , @xmath199 by theorem [ teo4.2 ] , the minimax risk among all predictive density estimators is asymptotically equivalent to the minimax risk among _ linear _ density estimators .
furthermore , by theorem [ teo3.2 ] , @xmath200 note that this minimax risk is strictly smaller than the minimax risk over the class of plug - in estimators since , for any plug - in density @xmath201 , @xmath202 = \frac{1}{2 } e \|\hat\theta - \theta\|^2\end{aligned}\ ] ] and by pinsker s theorem , the minimax risk of estimating @xmath44 under squared error loss is @xmath203 , which is larger than @xmath204 , by the fact that @xmath205 for any @xmath206 .
[ ex2 ] suppose that @xmath192 and @xmath44 is restricted in a sobolev ellipsoid @xmath207 where @xmath208 for @xmath209 then , by ( [ def : n ] ) , we have @xmath210 as @xmath162 . substituting this relation into equation ( [ def : tl ] ) yields @xmath211 using the taylor expression @xmath212 and the asymptotic relation @xmath213
we obtain @xmath214 where @xmath215^{{1}/{(2\alpha+1)}}.\ ] ] note that , by ( [ def : tth ] ) , @xmath216_+ \bigl(1+\mathrm{o}(1)\bigr).\end{aligned}\ ] ] therefore , @xmath217 by theorem [ teo4.2 ] , the minimax risk among all predictive density estimators is asymptotically equivalent to the minimax risk among the _ linear _ density estimators
. furthermore , by theorem [ teo3.2 ] , @xmath218 it is difficult to calculate an explicit form of the optimal constant for the minimax risk due to the @xmath219 function , but we can get an accurate bound for it . by taylor expansion , there exists @xmath220 , such that @xmath221 moreover , @xmath222 where @xmath223 therefore , @xmath224 that is , the convergence rate is @xmath225 and the convergence constant is between @xmath226 and @xmath227 . as in example
[ ex1 ] , we compare the asymptotics of this minimax risk with the one over the class of plug - in estimators , where the latter can be easily computed by ( [ plugin_risk ] ) and the results in @xcite .
direct comparison reveals that the convergence rates of both minimax risks are @xmath228 and the convergence constants can both be written in the form @xmath229 , where @xmath230 is a function depending only on @xmath231 .
although it is hard to obtain an explicit representation for the convergence constant for the overall minimax risk , our simulation result in figure [ fig1 ] shows that it is strictly smaller than that over the class of plug - in estimators . and @xmath232 .
] [ append ]
in this appendix , we provide the proofs of theorem [ teo2.1 ] and lemma [ lem4.1 ] .
proof of theorem [ teo2.1 ] let @xmath233 be an @xmath234 matrix whose @xmath235th entry equals @xmath236 . since the @xmath237 s form an orthogonal basis for @xmath8 and the @xmath238 s are equally spaced , we have @xmath239 .
consider the transformation @xmath240 .
since the first @xmath157 columns of @xmath233 are @xmath61 , the first @xmath157 elements of the transformed vector are just @xmath75 , defined in ( [ eq : transform ] ) , and we denote the remaining @xmath241 elements by @xmath242 . it is easy to check that @xmath243 and @xmath244 are independent multivariate gaussian variables , and the target density function @xmath15 satisfies @xmath245 where @xmath246 is the jacobian for this transformation .
similarly , any predictor density estimator @xmath20 can be rewritten as @xmath247 where @xmath74 is a transformation of @xmath248 defined in ( [ eq : transform ] ) .
note that the two predictive density functions on the left and right sides of the above equation may have different functional forms ; however , to simplify the notation , we use the same symbol @xmath36 to represent them when the context is clear .
now , the average kl risk can be represented as @xmath249\\[-8pt ] & = & { e}_{x , \x , \z | \th } \log \frac{p(\x , \z { \vert } \th)}{\hp(\x , \z { \vert } x)},\nonumber\end{aligned}\ ] ] where the second equality follows from ( [ trans_1 ] ) and ( [ trans_2 ] ) . since @xmath75 and @xmath242 are independent , we can split @xmath250 as @xmath251 where @xmath252 has a known distribution @xmath253 moreover , to evaluate the minimax risk , it suffices to consider predictive density estimators in the form @xmath254 because any predictive density @xmath255 can be written as @xmath256 , and if @xmath257 is equal to @xmath252 , then this density estimator is dominated by @xmath258 , due to the non - negativity of kl divergence . combining ( [ teo2.1:eq1])([teo2.1:eq3 ] ) , we have @xmath259 consequently , the minimax risk in the non - parametric regression model is equal to the minimax risk in the gaussian sequence model .
proof of lemma [ lem4.1 ] let @xmath260 be the collection of all ( generalized ) bayes predictive densities . then , by @xcite , theorem 5 , @xmath260 is a complete class for the problem of predictive density estimation under kl loss .
therefore , the minimax risk among all possible density estimators is equivalent to the minimax risk among ( generalized ) bayes estimators , namely , @xmath261 consider a gaussian distribution @xmath262 , where @xmath263 and the @xmath264 s satisfy condition ( [ cond : s_i ] ) .
then , @xmath265 the first term of ( [ ineq : bound_1 ] ) is the bayes risk under @xmath110 over the unconstrained parameter space @xmath37 .
it is achieved by the linear predictive density @xmath97 ; see @xcite .
therefore , @xmath266\\[-8pt ] & = & \frac{n}{2 m } \log \frac{\s2_n}{\s2_{n+m } } + \frac{1}{2m}\sum_{i=1}^n \log \frac{\s2_{n+m } + s_i^2}{\s2_n + s_i^2}.\nonumber\end{aligned}\ ] ] to bound the second term of ( [ ineq : bound_1 ] ) , note that for any bayes predictive density @xmath267 , @xmath268 where ( [ ineq : risk : bayes:1 ] ) is due to jensen s inequality , ( [ ineq : risk : bayes:2 ] ) is due to @xmath269 and ( [ ineq : risk : bayes:3 ] ) is due to @xmath270 therefore , @xmath271,\ ] ] where @xmath272 . using the cauchy
schwarz inequality , we can further bound the right - hand side of ( [ eq : bound : term2 ] ) as follows : @xmath273 \\ & & \quad \le \frac{1}{m v_m } \biggl [ \sum_{i=1}^n \biggl ( \int_{\th^c } \th_i^4 \pi_s(\th ) \,\mathrm{d } \th \biggr)^{1/2 } \sqrt{\pi_s(\th^c ) } + \frac{c}{a_1 ^ 2 } \pi_s(\th^c ) \biggr ] \\ & & \quad = \frac{1}{m v_m } \biggl [ \sqrt{3 } \sqrt{\pi_s(\th^c ) } \sum_{i=1}^n s_i^2 + \frac{c}{a_1 ^ 2 } \pi_s(\th^c ) \biggr ] \\ & & \quad \le \frac{1}{m v_m } \biggl [ \sqrt{3}\frac{c}{a_1 } \sqrt{\pi_s(\th^c ) } + \frac{c}{a_1 } \pi_s(\th^c ) \biggr].\end{aligned}\ ] ] then , by @xcite , proposition 2 , which states that if @xmath274 are independent gaussian random variables with @xmath275 and @xmath276 , then @xmath277 we have @xmath278^{1/2 } \le \s2_n^ { \alpha},\end{aligned}\ ] ] due to condition ( [ cond : s_i ] ) . combining ( [ ineq : bound_1 ] ) , ( [ ineq : term_1 ] ) , ( [ eq : bound : term2 ] ) and ( [ order ] ) , the theorem then follows immediately .
the authors would like to thank edward i. george for helpful discussions and the associate editor for generous insights and suggestions .
this work was supported in part by the national science foundation under award numbers dms-07 - 32276 and dms-09 - 07070 .
any opinions , findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the national science foundation . | we consider the problem of estimating the predictive density of future observations from a non - parametric regression model .
the density estimators are evaluated under kullback
leibler divergence and our focus is on establishing the exact asymptotics of minimax risk in the case of gaussian errors .
we derive the convergence rate and constant for minimax risk among bayesian predictive densities under gaussian priors and we show that this minimax risk is asymptotically equivalent to that among all density estimators . |
intense effort is being invested to improve our understanding of nuclear dynamics , e.g. the evolution of colliding nuclear systems and the fission process .
most successfully used theoretical approaches consider these processes in terms of a few macroscopic collective variables , representing the shape of the system , and the individual motion of non - interacting nucleons , describing the intrinsic excitation .
the dynamics of the system is described by transport equations of the fokker - planck or the langevin type in a non - equilibrium statistical approach .
a fundamental nuclear property , appearing as a parameter in these models , is the viscosity , respectively the dissipation strength , which measures the magnitude of the coupling between intrinsic and collective motion .
this key property in particular governs the relaxation phenomena and diffusion processes of the nuclear system .
dissipation plays an essential role in the calculation of the formation cross sections of super - heavy nuclei ( e.g. @xcite ) and in a consistent description of the fission process ( e.g. @xcite ) .
still , the magnitude of the dissipation strength and in particular an eventual dependence on temperature and/or deformation is under debate .
different theoretical models ( one - body dissipation : @xcite , linear response : @xcite ) yield diverging results .
moreover , quantum - mechanical effects , which go beyond these models , enter more and more into the discussion ( e.g. @xcite ) .
in this situation , it is of prime importance to improve our empirical knowledge in this field by well adapted experimental approaches .
the fission process has probably been the most successful tool for investigating nuclear dissipation .
it is characterized by a clearly observable large - scale collective motion , ending up in a binary decay of the system , which is in competition with statistical decay by sequential evaporation of nucleons and light clusters .
theoretically , one expects several phenomena to appear in fission due to nuclear viscosity : firstly , the system might show relaxation effects while establishing the quasi - equilibrium distribution in the shape coordinates .
this phenomenon was first studied by grang , jung - qing and weidenmller @xcite on the basis of the fokker - planck equation . as a consequence ,
the onset of the flow over the fission barrier is delayed by the transient time .
the system may then cool down by evaporation , reducing the fission probability .
transient effects are expected to manifest only under specific experimental conditions @xcite . secondly , when quasi - equilibrium is established , the quasi - stationary flow over the fission barrier is reduced with respect to the prediction of the bohr - wheeler transition - state model @xcite .
this reduction has already been deduced by kramers @xcite , who derived an analytical solution of the fokker - planck equation for an inverted parabolic potential .
these two phenomena , which influence the flow over the fission barrier , will be studied in the present work .
thirdly , another consequence of dissipation is an increase of the dynamical saddle - to - scission time in the quasi - stationary flow . from the experimental side ,
there exist a number of observations that support the dynamical nature of the fission process .
the anomalously enhanced pre - scission neutron multiplicities in fission induced in heavy - ion collisions @xcite have been interpreted as a signature of the delay of fission at high excitation energies .
other evidences for dissipative effects were obtained from the analysis of gamma - rays emitted during the de - excitation of the gdr @xcite or directly measuring the fission time using crystal blocking techniques @xcite . from these experiments ,
the increase of the dynamical saddle - to - scission time due to dissipative effects appears to be a well - established phenomenon .
the experimental proof for dissipative effects on the flow over the fission barrier remains an important subject of intensive research .
in contrast to heavy - ion fusion - fission reactions , which suffer from complex initial conditions , e.g. broad angular - momentum distributions and large shape distortions , nucleon - induced reactions or very peripheral nucleus - nucleus collisions at relativistic energies seem to be best suited for these studies . recently , a novel experimental approach has been introduced @xcite , which is based on the detection of fission products from nucleon - induced reactions or very peripheral nucleus - nucleus collisions at relativistic energies in inverse kinematics .
the advantage of these reaction mechanisms is that the excited fissioning nucleus is produced with well defined initial conditions . in particular , these reactions lead to almost undeformed nuclei . in addition , they cover a large range in excitation energy while introducing only moderate angular momenta @xcite . in these experiments , reaction residues with nearly projectile velocity are detected in forward direction .
the inverse - kinematics conditions allow for identifying the fission products in atomic and mass number .
these experiments were designed to be particularly sensitive to the time needed by the highly excited system to reach quasi - equilibrium . in these works ,
the fission cross sections , the charge distribution @xcite , and the isotopic distribution @xcite of fission residues have been used as signatures of the fission dynamics .
evidences for relaxation effects leading to transient times in the order of a few 10@xmath1 s were deduced @xcite , which corresponds to conditions close to the minimun possible transient time realized at critical damping @xcite . since several years , also experiments on light charged
- particle - induced reactions , e.g. @xcite in conventional kinematics have been performed . in this case ,
evaporative neutrons and light charged particles were measured , from which initial thermal excitation energies were evaluated event by event .
in addition , the fission fragments were registered , however , with a rather low resolution in mass . in a very recent letter @xcite ,
the measured fission probabilities of highly excited systems in 2.5-gev proton - induced reactions were compared with model calculations . from this comparison
the authors could not find any indication for dynamical effects on the flux over the fission barrier , since the data were fully explained by the statistical bohr - wheeler approach .
thus , the findings deduced from the signatures of these two different experiments are in clear contradiction .
in addition , the conclusions of ref .
@xcite are rather intriguing by themselves .
in particular , the following two statements of the paper are hardly compatible in view or our present theoretical understanding of the fission process : on the one hand , the protons are expected to induce only little collective excitations by angular momentum , shape distributions and compression . on the other hand , it is claimed that fission is in competition with particle evaporation immediately after the nuclear - collision stage of the reaction .
it seems that the deduced immediate onset of fission requires a very fast population of the transition states above the fission barrier in a time well below the minimum transient time imaginable .
thus , the conclusions drawn in ref . @xcite seem to require a revised understanding of nuclear dynamics . if we admit that the inertia of the nuclear system is rather well established , it seems that the validity of the non - equilibrium statistical approach in terms of the transport equations of the fokker - planck or the langevin type must be questioned .
the present work is intended to contribute to the solution of these problems .
we investigate the reaction @xmath0u on deuterium at 1 a gev in inverse kinematics , taking advantage of the unique experimental installations of gsi .
however , this time we use a different observable , the isotopic production cross sections of evaporation residues . due to the high fissility of the projectile , the yields of evaporation residues are expected to strongly depend on the fission process and in particular on transient and quasi - stationary dissipative effects on the fission flux across the barrier .
in addition , the large excitation - energy range covered in the collisions induced by @xmath0u on deuterium can be sorted according to the mass loss of the final residues with respect to the projectile . because of the narrow mass range produced in the collision stage of the reaction , the mass loss is dominated by the evaporation process , which is directly related to the excitation energy induced in the collision .
one can then investigate the influence of transient and quasi - stationary dissipative effects in the fission flux as a function of the excitation energy involved in the fission process .
thus , the information on the physics of the fission process provided by the present experiment is very similar to the one obtained in ref .
@xcite , although it is based on different observables .
in addition , the system is very similar , providing almost the same amount of centre - of - mass energy .
the experiment was performed at gsi ( darmstadt ) where the sis synchrotron was used to produce a pulsed beam of @xmath0u at 1 a gev with a maximum intensity of 10@xmath2 ions s@xmath3 , a pulse length of 3 s and a repetition cycle of 7 s. this beam impinged onto a liquid - deuterium target with a thickness of 200 mg/@xmath4 .
the cylindrical target container was 1 cm long and 3 cm in diameter with titanium windows of 18 mg/@xmath4 in total . due to the inverse kinematics ,
the projectile residues produced in the reaction were emitted in forward direction and could be magnetically analysed in - flight with the fragment separator frs @xcite .
this is an achromatic two - stage zero - degree spectrometer with a dispersive intermediate image plane .
every stage is equipped with two dipoles and a set of quadrupoles in front and behind each dipole .
the resolving power of this device is b@xmath5/@xmath6b@xmath5 @xmath7 1500 with an acceptance of 15 mrad around the beam axis in angle and @xmath81.5% in momentum . in order to isotopically identify all the transmitted nuclei , their horizontal positions at the intermediate and final image planes
, their velocity and their energy loss in a given gas volume were measured with different detectors .
two plastic scintillators , about 20 cm long , 8 cm height and 5 mm thick , equipped with two photomultipliers at both extremes , were placed at the intermediate and final image planes to provide the position and time - of - flight measurements @xcite .
the atomic number of the nuclei was obtained from their energy - loss measurement in two multi - sampling ionisation chambers @xcite located behind the spectrometer .
finally , a set of multi - wire chambers placed at the different image planes of the spectrometer was used for calibration at the beginning of the experiment .
in addition , and in order to identify the atomic number of heavy reaction residues with z above 70 , a velocity degrader @xcite at the intermediate image plane of the spectrometer was used . the final resolution ( fwhm )
achieved for the mass and charge separation was a/@xmath6a @xmath7 400 and @xmath6z @xmath7 0.4 , respectively .
the mass and charge identifications of the reaction residues were obtained by using the primary beam as reference and the specific pattern of light nuclei when represented on a two - dimensional spectrum a / z versus z @xcite . due to the limited acceptance of the spectrometer ,
more than 50 different magnetic tunings of the dipoles of the frs were needed to completely measure the momentum distributions of all nuclear species produced in the reaction @xmath0u on deuterium at 1 a gev with a production cross section above 100 @xmath9b and z@xmath1023 . using the information provided by the beam - monitor detector seetram @xcite , the yields measured in different magnetic tunings were normalised and combined to produce the corresponding momentum distribution for each nuclide .
the integral of these momentum distributions normalised to the number of atoms in the target and corrected by the detection efficiency , angular transmission and secondary reactions provided the production cross sections of all measured nuclei . a detailed description of the experiment and the data analysis for a similar experiment can be found in references @xcite .
the results of the experiment are summarised in fig.1 . in this figure
, we represent in a two - dimensional cluster plot all the projectile residues produced in the reaction @xmath0u on deuterium at 1 a gev , with a cross section larger than 100 @xmath9b and z above 23 , on top of a chart of the nuclides .
more than 1400 different nuclides have been identified , and their cross sections have been determined with an accuracy between 10% and 20% . in the figure we can clearly observe two different groups of nuclides .
the upper part of the chart of the nuclides is populated by mostly neutron - deficient nuclides covering a large range in charge from the projectile nucleus down to the charge 65 . on average , for a given element , the isotopic distribution covers around 20 different nuclides with a maximum production cross section around a few millibarns .
these nuclei are produced in a nucleon and/or cluster evaporation process from the excited projectile pre - fragments produced in the interaction between the projectile and the target nuclei .
this process leads to residual nuclei lighter than the projectile with an isotopic composition determined by the competition between neutron and proton evaporation and the initial excitation energy induced in the collision . since the evaporation of protons and neutrons
is governed not only by the respective binding energies but also by the proton coulomb barrier , the equilibrium between both processes is reached at the left of the stability line , along the so called `` evaporation corridor '' @xcite , defining the ridge line of the nuclide distributions of evaporation residues .
the initial excitation energy determines the length of the evaporation chain and consequently the mass of the final residue .
since the number of nucleons removed from the projectile in direct collisions with the target deuteron is expected to be rather small , the mass of the evaporation residues is an observable which is well correlated with the excitation energy induced in the reaction .
this important aspect will be discussed later in further detail .
the second group of nuclei observed in fig .
1 corresponds to medium - mass fragments with atomic numbers between 23 and 65 . in this case , the isotopic distributions are broader , populating on average 25 isotopes .
the distributions are centred to the right of the stability line , and the production cross sections are much larger than the ones observed for the evaporation residues .
these nuclei are expected to be produced by fission of the projectile pre - fragments emerging from the interaction between the @xmath0u projectiles with deuterium .
the neutron excess of the projectile pre - fragments is preserved in the fission process . however , when the system crosses the fission saddle at high excitation energies , this neutron excess is partially lost by post - saddle neutron evaporation , both from the system on the way to scission and from the fragments , shifting and broadening the final isotopic distribution of fission residues .
the observed mass distribution of fission residues is mostly single - humped , in contrast to the well known double - humped distribution of residues produced in the low - energy fission of @xmath0u , indicating in our case that fission takes place at high excitation energy , where shell effects are washed out .
nevertheless , the most neutron - rich fission residues show a contribution by a double - humped component resulting from low - energy fission induced in very peripheral reactions .
u(1 a gev ) ( black dots ) and @xmath11pb(1 a gev ) ( grey dots ) @xcite on deuterium as a function of the mass loss of the final residues with respect to the mass of the projectile .
the isobaric distribution of residues produced in the reaction @xmath11pb(1 a gev ) on deuterium is compared to a model calculation ( solid line ) performed with the isabel intra - nuclear cascade code @xcite coupled to the abla evaporation code @xcite including a dynamical description of fission according to reference @xcite.,width=340 ] from the analysis of fig .
1 we can conclude that fission is the dominant de - excitation channel in the reaction @xmath0u on deuterium at 1 a gev . a similar conclusion
can be drawn from fig .
2 . in this figure
, we represent the isobaric distribution of projectile residues measured in the reaction @xmath0u(1 a gev ) on deuterium ( black dots ) compared to the one obtained for the reaction @xmath11pb(1 a gev ) on deuterium ( grey dots ) @xcite .
we can also identify the two groups of residues produced in evaporation and fission processes , respectively .
although the excitation energy induced in both reactions should be similar , fission seems to be the dominant reaction channel in the collisions induced by @xmath0u , while in the case of the reactions induced by @xmath11pb we arrive to the opposite conclusion .
this result can be attributed to the different fissilities of the pre - fragments produced in both reactions .
since the fission component for the @xmath11pb system represents less than 15% of the total cross section , and the total cross sections for the reactions @xmath0u(1 a gev ) and @xmath11pd(1 a gev ) on deuterium differs by less than 15% , we can conclude that the differences observed in fig .
2 for the isobaric distributions of evaporation residues in both reactions are mostly due to the fission process in the reactions induced by @xmath0u at 1 a gev on deuterium .
a deeper and more quantitative interpretation of our data requires the use of model calculations . in order to describe the interaction of deuterium with @xmath0u at relativistic energies , we used an intra - nuclear cascade code coupled to a de - excitation code considering both nucleon and/or cluster evaporation and fission .
after benchmarking different intra - nuclear - cascade models , the most consistent description of the available data on @xmath0u + @xmath12h and @xmath11pb + @xmath12h @xcite , which are characterised by substantially different fissilities , was obtained with the isabel code of yariv and fraenkel @xcite .
the evaporation and fission processes were described using the monte - carlo code abla . in this code
the evaporation of nucleons and alpha clusters is described according to the weisskopf formalism including a consistent description of level densities where shell , pairing and collective effects are considered following reference @xcite .
the fission width is calculated using the statistical approach of bohr and wheeler @xmath13 @xcite , corrected for quasi - stationary and transient dissipative effects , which describe the dynamical nature of fission according to : @xmath14 in this equation @xmath15 represents to the so - called kramers coefficient which provides the correction one has to apply to the statistical fission width in order to obtain the quasi - stationary solution of the fokker - plack equation describing the fission flux across the barrier @xcite .
the function @xmath16 describes the time dependence of the fission width resulting from the time - dependent solution of the fokker - planck equation proposed by grang , jun - qing and weidenmller @xcite . the analytical approach of ref .
@xcite to eq .
[ eq_1 ] makes it possible to avoid the computational effort required to solve numerically the fokker - planck equation on a event - by - event basis . in eq .
[ eq_1 ] the kramers factor is defined as : @xmath17^{1/2 } - \frac{\beta}{2\omega_o } \right\ } \label{eq_2}\ ] ] where @xmath18 is the reduced dissipation coefficient , and @xmath19 corresponds to the frequency of the harmonic oscillator describing the inverted potential at the fission barrier . in our monte - carlo code
we use different analytical approximations to describe the time dependence of the fission width @xmath16 according to ref .
@xcite . the main consequence of eq . [ eq_2 ]
is that the quasi - stationary and transient effects that appear in the dynamical description of the fission flux across the barrier lead to a reduction of the fission width compared to the value obtained with the statistical model . besides nuclear dissipation , the fission barriers and even more the level densities are the ingredients of the model description , which have the most important influence on the fission probabilities at high excitation energies and thus should be considered with special care .
in addition to the arguments given in @xcite , which already justified the choice of ref .
@xcite and @xcite for the description of fission barriers and level densities , the deformation dependence of the level - density parameter , which is the basis for their ratio at the transition state and ground deformation a@xmath20/a@xmath21 used in our calculation , has been investigated recently on the basis of the yukawa - plus - exponential description of the nuclear properties @xcite .
also this theoretical study has essentially confirmed the validity of ref .
@xcite , which we use in our calculation .
finally , the isotopic distribution of fission residues is described following the model of reference @xcite .
in addition , the abla code includes a breakup channel which sets in when the nuclear temperatures reaches 5 mev @xcite .
an example of the results obtained with this model is shown in fig .
2 ( solid line ) for the reaction @xmath11pb(1 a gev ) on deuterium .
this reaction can be considered as an optimum case to benchmark the intra - nuclear and the evaporation code , since in this case the role of fission is minor .
moreover , the description of the level densities is simpler since only few nuclei will feel the z=82 shell , and the collective enhancement for the deformed nuclei is expected to be almost the same for the different de - excitation channels , e.g. in the ground state for particle emission and at the saddle for fission . in this calculation we described the fission dynamics by means of the analytical approximated solution of the time - dependent fokker - planck equation , proposed by b. jurado and collaborators @xcite , with a reduced dissipation coefficient @xmath22 .
the good description of the production cross sections provided by the code for both , evaporation and fission residues , validates the intra - nuclear cascade model as well as the description of the evaporation channels and the breakup , which plays a role only for the lightest evaporation residues . in order to better understand the general trends of the reaction @xmath0u(1a gev)+d
, we show in fig .
3 the calculated excitation energy of the initial pre - fragments as a function of the mass - loss of the final residue ( @xmath6a ) with respect to the projectile .
this calculation was done with the same description of the fission width used for the calculation of fig . 2 .
u ( 1a gev)+d.,width=302 ] as can be seen , only evaporation residues ( @xmath6a @xmath23 90 ) exhibit a clear correlation between their final masses and the initial excitation energies of the decaying pre - fragments . applying different intra - nuclear cascade codes @xcite
, we found that this correlation is a common feature , which quantitatively varies only little for the different codes .
we understand this result as a consequence of a rather strict correlation between the initial excitation energy and the mass loss in the de - excitation stage , the mass loss in the intra - nuclear cascade stage being comparably small . in the light of this figure
, we can re - examine the residue productions shown in fig . 2 for the reaction @xmath0u ( 1a gev)+d .
considering the existing correlation between the measured mass loss of the final evaporation residues with respect to the projectile and the initial excitation energy of the pre - fragments and the differences in the yields of evaporation residues in collisions induced by @xmath0u and @xmath11pb , we arrive to an interesting observation : the fission channel seems to dominate the de - excitation process of pre - fragments produced in the reaction @xmath0u(1 a gev ) on deuterium at low and moderate excitation energies .
( note that fig .
2 provides an energy scale to the ordinate of fig .
3 ) . however , the similar production cross sections of light evaporation residues ( @xmath6a @xmath7 70 ) observed in both reactions clearly indicate that the fission of @xmath0u pre - fragments is suppressed at high excitation energies .
this is a surprising result that clearly contradicts the expectations due to the statistical description of the fission process according to the model of bohr and wheeler @xcite .
this inhibition of fission at high excitation energies would qualitatively fit with the dynamical interpretation of fission as proposed by kramers @xcite and later on by grang , jun - qing and weidenmller @xcite .
u(1 a gev)+d ( dots ) , compared with different calculations performed with the isabel intra - nuclear cascade code coupled to the abla evaporation code ( see text for details).,height=264 ] once we have validated most of the ingredients in our code we can concentrate on the fission channel which is the dominant one in the reaction induced by @xmath0u . in fig .
4 , we compare the measured isobaric distribution of evaporation residues produced in the reaction @xmath0u(1 a gev)+d ( dots ) with different model calculations .
the dotted line was obtained using the purely statistical model of bohr and wheeler @xcite . as can be seen
, this model clearly overestimates the fission channel depopulating the production of evaporation residues . better agreement with the measured data
is obtained when the dynamics of fission is considered in the calculation .
the dashed line corresponds to a calculation where the fission width is evaluated using the kramers factor described by equation [ eq_2 ] with a reduced viscosity parameter of @xmath18 = 2@xmath2410@xmath25 s@xmath3 .
however , a calculation considering not only the stationary effects in the fission flux by means of the kramers factor with @xmath18 = 2@xmath2410@xmath25 s@xmath3 , but also the transient effects due to the time dependence of the fission width ( solid line ) provides the best description of the data . in this case
, the time dependence of the fission width has been calculated according to the approximated solution of the fokker - planck equation proposed by jurado and collaborators @xcite . the fair agreement observed in fig .
4 between our calculations ( solid line ) and the experimental data ( dots ) is consistent with previous investigations where different observables and reactions have been used . in particular , partial fission cross sections and
final charge distributions of fission residues produced in the reaction @xmath0u(1 a gev)+ch@xmath26 @xcite , total fission cross section and the isotopic distribution of fission residues in collisions induced by @xmath27au on hydrogen at 800 a mev @xcite and the present isotopic distributions of evaporation residues produced in the reaction @xmath0u at 1 a gev on deuterium , are coherently reproduced by these calculations . however , until now , none of the observables used as signatures of the fission dynamics allowed us to characterise and quantify independently the role of transient and quasi - stationary dissipative effects in the fission flux at small deformation . in principle
one would expect low - energy fission to be sensitive to quasi - stationary effects , characterised by the reduced dissipation coefficient in kramers factor , and not to transient effects , which are expected to be much shorter than the statistical fission time at low excitation energy .
on the other hand , as the excitation energy increases , the statistical fission time approaches the transient time , and the fission process becomes sensitive to this latter effect .
the strong correlation between the final mass of the evaporation residues and the initial excitation energy , together with the large excitation - energy range covered with the reaction @xmath0u at 1 a gev on deuterium allowed us to analyse our data following these ideas . in fig .
5 we show the isobaric distribution of evaporation residues produced in the reaction @xmath0u at 1 a gev on deuterium compared with calculations using different reduced dissipation coefficients and descriptions of the time - dependent fission width . in the left panel of the figure we compare the measured data to calculations , where different values of the reduced dissipation coefficient @xmath18 have been used , @xmath18 = 1@xmath2410@xmath25 s@xmath3(dashed line ) , @xmath18 = 2@xmath2410@xmath25 s@xmath3(solid line ) , @xmath18 = 3@xmath2410@xmath25 s@xmath3 ( dotted line ) and @xmath18 = 5@xmath2410@xmath25 s@xmath3 ( dotted - dashed line ) . in all these calculations
the time dependence of the fission width has been described using the approximate solution of the fokker - planck equation proposed by jurado et al .
these calculations clearly show how the measured yields of evaporation residues depend on the value of the reduced dissipation coefficient .
even more , this effect manifests mainly in heavy evaporation residues corresponding to the survival probability against fission at low or moderate excitations energies .
as already stated , the value of the reduced dissipation coefficient @xmath18 = 2@xmath2410@xmath25 s@xmath3 reproduces the data fairly well . in the right panel of fig .
5 , we compare the same data to another set of model calculations .
this time we have fixed the value of the reduced dissipation coefficient to @xmath18 = 2@xmath2410@xmath25 s@xmath3 and we have used different models to describe the time dependence of the fission width . the result of the calculation using the approximate solution of the fokker - planck equation proposed by jurado et al .
@xcite is represented by the solid line .
the dotted line corresponds to an exponential in - growth time dependence of the fission width according to the expression @xcite @xmath28 where @xmath15 is the kramers factor , @xmath13 the statistical fission width and @xmath29=@xmath30/2.3 with @xmath30 being the transient time defined as 90% of the time the fissioning system needs to reach the stationary fission - decay width across the barrier . in this figure
, the dashed line represents the result obtained with a time dependence of the fission width following a step function @xcite .
@xmath31 from this comparison we can conclude that the time dependence of the fission width based on the approximated solution of the fokker - planck equation and the step function provide similar results , although the first one has a better physical justification . the exponential in - growth clearly overestimates the fission width , in particular at high excitation energies ( short times ) as was already pointed out in reference @xcite .
in addition , we can also conclude that in contrast to the calculation performed with different values of the reduced dissipation coefficient @xmath18 , in this case the sensitivity of the data to the transient effects appears for large values of the mass loss of the residues corresponding to high excitation energies .
consequently , our data make it possible to decouple the role of transient and quasi - stationary dissipative effects in the fission flux .
we describe the evaporation - residue cross sections for a large mass range with the same value of the reduced dissipation coefficient ( @xmath18 = 2@xmath2410@xmath25 s@xmath3 ) fairly well .
slight deviations for the lightest residues might indicate an even stronger suppression of fission at the highest excitation energies , since the measured cross sections of these residues are slightly underestimated by the calculation . before drawing more quantitative conclusions , improved theoretical calculations are planned to better describe the entrance - channel distribution of the pre - fragments in deformation space , which we assumed to be identical to the zero - point motion of the projectile around a spherical shape in the present analysis .
when we compare our conclusions with the ones drawn in ref .
@xcite , we observe a severe discrepancy .
while we find strong indications for dissipative effects in the magnitude of the evaporation - residue cross sections , the essentially complementary fission probabilities measured in @xcite were fully reproduced by calculations with the purely statistical bohr - wheeler approach . for the system
@xmath0u + @xmath12h , we noticed an essential difference in the ingredients of the model calculation , which probably explains the controversial conclusions : while we determine the appropriate value of a@xmath20/a@xmath21 according to the saddle - point deformation of each fissioning system following the description of ignatyuk et al .
@xcite for the deformation dependence of the level density parameter , tishchenko et al .
@xcite use a common value of a@xmath20/a@xmath21 = 1 for all fissioning systems produced in the collision of @xmath0u with 2.5 gev protons , which extend over a broad range of elements @xcite .
with the option a@xmath20/a@xmath21 = 1 , we also can reproduce the fission probabilities reported in ref .
@xcite within the bohr - wheeler statistical approach fairly well without introducing any transient effects .
thus , the diverging findings of tishschenko et al . and the present work for the fission of @xmath0u by 2 gev light charged particles seem to be traced back to different ingredients of the model calculations .
a discussion of the results for the lighter systems also studied in @xcite is beyond the scope of the present work .
we will address this interesting subject in the near future , when data from inverse - kinematics experiments for similar systems will be available .
in this letter , we have shown that the accurate measurement of the isotopic production of reaction residues provides valuable information on the reaction mechanism . in particular , we have presented the measured isotopic cross sections of more than 1400 different projectile residues produced in the reaction @xmath0u(1 a gev ) on deuterium with an accuracy between 10% and 20% . in this reaction
, the dominant fission channel can be investigated from the measured fission residues or from the survival probability against fission , represented by the evaporation residues . however , the strong correlation existing between the mass loss of the evaporation residues and the initial excitation energy induced in the collisions offers the possibility to use these residual nuclei for investigating the excitation - energy dependence of the fission process .
the comparison of these data with the evaporation residues produced in the reaction @xmath11pb(1 a gev ) on deuterium clearly shows the dominance of fission in the @xmath0u - induced reaction .
however , the fact that for large mass losses both reactions lead to the production of residues with similar cross sections represent a clear signature of the suppression of fission in the @xmath0u - induced reactions at high excitation energies .
model calculations show that the measured data can only be understood when the dynamics of fission is considered , in agreement with previous works . however , the accuracy of the present data , together with the large excitation - energy range covered with this reaction made it possible to characterise and quantify independently the role of transient and quasi - stationary dissipative effects in the fission flux across the barrier .
the cross sections in the whole mass range are fairly well reproduced with a single value of the dissipation coefficient @xmath18 = 2@xmath2410@xmath25 s@xmath3 .
discrepancies in the interpretation of similar data recently obtained by tishchenko el al .
@xcite in direct kinematics were understood in terms of the different ingredients used in the model calculations .
the use of appropiate values for the level density parameter and in particular its dependence with deformation seems to be crucial for the correct understanding of the fission dynamics at high excitation energy .
we thank b. jurado , a. kelic , f. rejmund and c. volant for valuable discussions as well as t. enqvist for the careful reading of the manuscript .
this work was partially supported by the european commission through the hindas project under contract fikw - ct-2000 - 031 and the spanish mec and xuga under contracts fpa2002 - 04181-c04 - 01 and pgidt01pxi20603pm , respectively . | isotopic cross sections of all projectile residues with z above 23 produced in collisions induced by @xmath0u at 1 a gev on deuterium have been measured .
the isotopic distributions reflect the role of evaporation and fission in the formation process of these nuclei .
the comparison of the measured cross sections with monte - carlo de - excitation codes including an analytical description of the dynamics of fission shows the sensitivity of the data to nuclear dissipation .
moreover , the large excitation - energy range covered in this experiment together with the high accuracy of the measured cross sections allowed to clearly separate and quantify the role of transient and quasi - stationary dissipative effects in the fission - decay width . |
significant fraction of isolated white dwarfs form with strong magnetic fields .
these high - field magnetic white dwarfs ( hfmwd ) , the majority of which were discovered via the sloan digital sky survey ( sdss ; @xcite ) , comprise @xmath010% of all isolated white dwarfs @xcite .
surface field strengths range from a few to slightly less than a thousand megagauss ( mg ) while the bulk of the isolated wd population have measured weak fields or non - detection upper limits of typically @xmath1 g @xcite . in wd - binary systems ,
the companion s surface is an equipotential . in tight binaries
, this equipotential extends toward the wd , leading to mass transfer from the companion onto the wd , a process called `` roche - lobe overflow . '' among systems such as these , a possibly even larger fraction of the wd primaries are highly magnetic ( i.e. , @xmath025% of cataclysmic variables ; @xcite ) . magnetic cataclysmic variables ( cvs ) are generally divided into two classes : am herculis ( polars ) and dq herculis ( intermediate polars ) . for reviews on
am her and dq her systems see @xcite and @xcite
. generally , am her systems are those in which both components of the binary are synchronously rotating at the orbital period . in this scenario ,
the formation of an accretion disk is prevented and material is funneled onto the wd via the magnetosphere .
polars have strong magnetic fields ( @xmath0@xmath2 g ) , copious x - ray emission and stable pulsations in their lightcurves .
on the other hand , dq her systems ( intermediate polars ) are not synchronously locked and have weaker magnetic fields than their polar counterparts , often by an order of magnitude or more . for these systems , an accretion disk forms from which the wd is spun up .
remarkably , _ not a single _ observed close , detached binary system ( in which the primary is a wd and the companion is a low - mass main - sequence star ) contains a hfmwd @xcite . if the magnetic field strengths of white dwarfs were independent of binary interactions , then the observed distribution of isolated wds should be similar to those in detached binaries .
in particular , within 20 pc , there are 109 known wds ( 21 of which have a non - degenerate companion ) , and sdss has identified 149 hfmwds ( _ none _ of which has a non - degenerate companion ) . assuming binomial statistics , the maximum probability of obtaining samples
at least this different from the same underlying population is @xmath3 , suggesting at the 6.2-@xmath4 level that the two populations are different ( for details on this kind of calculation , see appendix b of @xcite ) .
furthermore , sdss identified 1253 wd+m - dwarf binaries ( _ none _ of which are magnetic ) .
as was previously pointed out , this suggests that the presence or absence of binarity is crucial in influencing whether a hfmwd results @xcite .
these results initially seem to indicate that hfmwds preferentially form when isolated . however , unless there is a mechanism by which very distant companions prevent the formation of a strong magnetic field , a more natural explanation is that highly - magnetized white dwarfs became that way by engulfing ( and removing ) their companions . in the binary scenario ,
the progenitors of hfmwds are those systems that undergo a common envelope ( ce ) phase during post - ms evolution .
in particular , magnetic cvs may be the progeny of common envelope systems that almost merge but eject the envelope and produce a close binary .
subsequent orbital reduction via gravitational radiation and tidal forces turn detached systems into those which undergo roche - lobe overflow . for ces
in which the companion is not massive enough to eject the envelope and leave a tight post - ce binary , the companion is expected to merge with the core .
it was suggested that these systems may be the progenitors of isolated hfmwds @xcite . in this paper , we calculate the magnetic fields generated during the common envelope phase .
we focus on low - mass companions embedded in the envelope of a post - ms giant . during in - spiral
, the companion transfers orbital energy and angular momentum , resulting in differential rotation inside the ce . coupled with convection , a transient @xmath5 dynamo amplifies the magnetic field at the interface between the convective and radiative zones where the strongest shear is available .
the fields produced from this interface dynamo are transient and unlikely to reach the white dwarf surface with sufficient strength to explain hfmwd observations .
if , however , the companion tidally disrupts , the resultant super - eddington accretion disk can amplify magnetic fields via a disk dynamo .
the disk - generated fields are strong ( @xmath6 mg ) and accretion provides a natural mechanism with which to transport the fields to the wd surface . for a range of disrupted - companion masses ,
the fields generated in the disk are sufficient to explain the range of observed hfmwds .
common envelopes are often invoked to explain short period systems in which one component of the binary is a compact star @xcite .
the immersion of a companion in a ce with a post - main sequence star can occur via direct engulfment or through orbital decay due to tidal dissipation in the giant s envelope @xcite .
once engulfed , the companion in - spirals due to hydrodynamic drag until it either survives ( ejects the envelope leaving a post - ce close binary ) or is destroyed .
while ces can form for various mass - ratio binaries , we focus on low - mass stellar and substellar companions such that the remnant system is expected to be an isolated white dwarf .
since low - mass companions do not release enough orbital energy to eject the envelope , as the orbital separation is reduced , the differential gravitational force due to the proto - wd tidally shreds the companion .
the disrupted companion then forms a disk inside the ce that subsequently accretes onto the proto - wd @xcite . for more detail on the onset and dynamics of the ce phase for post - ms giants and low - mass companions ( i.e. low , mass - ratio binaries ) ,
see @xcite and @xcite .
as a consequence of common envelope evolution , the transfer of orbital energy and angular momentum during in - spiral generates strong shear . coupled with convection , shear leads to large - scale magnetic field amplification via an @xmath5 dynamo . during in - spiral , the free energy in differential rotation available to the dynamo is proportional to the companion mass . in general , the more massive the companion , the more free energy in differential rotation @xcite .
the dynamo converts free energy in differential rotation into magnetic energy and therefore , strong shear leads to strong magnetic fields .
some investigations of the dynamo in this context have imposed a velocity field to determine what steady - state magnetic field might arise if the velocity were steadily driven @xcite . however , as the magnetic field amplifies , differential rotation decreases . in general , for systems in which shear is not resupplied , a transient dynamo and decay of the magnetic field result @xcite .
this scenario was investigated as a way to generate the strong fields necessary to power bipolar outflows in post - asymptotic giant branch ( post - agb ) and planetary nebulae ( pne ) @xcite . a similar , but weaker , dynamo ( akin to the solar dynamo )
may operate in isolated red giant branch and asymptotic giant branch stars @xcite . here , we investigate two scenarios for magnetic field generation during a ce phase between a post - ms giant and a low - mass companion .
first , we calculate the magnetic fields at the interface between the convective and radiative zones in the ce ( hereafter referred to as the envelope dynamo scenario ) and show that this scenario , while potentially viable under special circumstances , is unlikely in general to explain the origin of hfmwds . as an alternative , we estimate the magnetic fields generated in the disk of a tidally disrupted companion around the proto - wd .
the disk - generated fields naturally accrete onto the wd surface and are sufficiently strong to match observations .
the shear profile in a giant star that is produced by a low - mass companion s ce - induced in - spiral leads to a dynamo in the presence of the turbulent convective zone .
the field primarily amplifies at @xmath7 , the interface between the convective and radiative regions .
the back - reaction of field - amplification on the shear is included with differential rotation and rotation depleting via loss of poynting flux and turbulent dissipation .
for the precise equations solved , generic features of envelope dynamos , and a pictorial representation of the dynamo geometry in isolated stellar and ce settings see @xcite . here
we calculate an upper limit to the magnetic field that may be generated by this mechanism , by taking the shear profile in the stellar envelope to be keplerian ( which produces stronger shear than could possibly be maintained in a spherical - hydrostatic star ) . before presenting details of these calculations , we note a few generic features of this type of scenario that might make the envelope dynamo an unlikely explanation of hfmwds , irrespective of details .
there are a few reasons why a field generated at the radiative - convective boundary may not be able to produce strongly magnetized material in the vicinity of the wd .
first , as shown below , a robust upper limit to the magnitude of the envelope dynamo generated field is a few times @xmath8 g. to explain megagauss fields at the wd surface , this requires an inwardly increasing gradient of @xmath9 ( e.g. via field amplification from flux freezing ) , which might lead the highest - field regions to rise buoyantly , therefore never reaching the wd surface .
second , since the envelope is transient ( operates for @xmath0100 years ) and the typical agb lifetime is @xmath0@xmath10 years , the envelope would need to be ejected at the time of formation of a hfmwd . without fine tuning , this results in a tight , post - ce binary with an under - massive white dwarf the exact opposite of the observed hfmwds . finally , even if the field could penetrate to the wd surface
, some mass would need to remain and/or fallback during envelope ejection to anchor the field .
a further subtlety is that the transport of magnetic flux into the radiative layer is likely to involve not merely isotropic diffusion but anisotropic diffusion , as a downward pumping of magnetic flux involves anisotropic convection .
we are led to what seems to be a more natural formation explanation , which is described in the next section .
nevertheless , despite the aforementioned potential difficulties , we now investigate the viability of the envelope dynamo scenario .
our stellar evolution models were computed using the `` evolve zero - age main sequence '' ( ez ) code @xcite.townsend / static.php?ref = ez - web ] we employ a zero - age main - sequence progenitor of 3 @xmath11 , with solar metallicity at the tip of the agb . under the assumption of keplerian rotation ,
the saturated toroidal field strength at the base of the 3-@xmath11 progenitor s convective zone ( @xmath12 cm ) is roughly @xmath13 g. to survive past the pn stage , the fields must reach the wd surface ( which is the core of the agb star , a distance @xmath14 interior to the base of the convective zone ; see fig . 3 of @xcite for a pictorial representation ) , anchor , and sustain or induce a field in the wd .
the latter scenario has been investigated by @xcite , who conclude that an ordered external field ( potentially generated from a dynamo ) can induce a surface field on the wd that decays to a few percent of its initial value after a million years . for the fields to reach the wd surface ,
radially inward diffusion of magnetic flux must act on a faster timescale than magnetic buoyancy , which transports flux outward .
the buoyant rise velocity , @xmath15 , is found by equating the upward buoyancy force on a flux tube to the downward viscous drag , thereby obtaining the upward terminal velocity .
it may be represented as @xmath16 , where @xmath17 is the alfvn velocity , @xmath18 is the convective fluid velocity , @xmath19 is the flux tube radius , @xmath20 is a pressure scale height , and @xmath21 is a dimensionless quantity of order unity @xcite . the time to buoyantly rise a distance @xmath22 from @xmath23 , assuming @xmath24 , is @xmath25 years .
the diffusion timescale to traverse a distance @xmath22 , given a turbulent diffusivity @xmath26 , is @xmath27 and must be less than the buoyant - rise time @xmath28 .
this implies a constraint on the turbulent diffusivity : @xmath29 . for @xmath30
cm , the buoyant - rise velocity of a @xmath31-g flux tube in an agb star with @xmath32 g @xmath33 , requires a diffusivity of at least a few times @xmath34 @xmath35 s@xmath36 ; a very large and physically unlikely value .
note however that if weaker fields are generated , the requirement on the diffusivity would lower correspondingly .
furthermore , during one cycle half - period @xmath37 ( defined as the time for one reversal of the field ) , magnetic flux of a given sign ( positive or negative ) can diffuse into the radiative layer .
years ) ; @xcite ) .
note that while the radiative layer is convectively stable it may still possess non - negligible turbulence . ] after reversal , field of the opposite sign amplifies and diffuses in the radiative layer .
therefore , the diffusion timescale must not be greater than the cycle half - period : @xmath38 .
this implies another a theoretical constraint on the turbulent diffusivity for consistency of our dynamo model , namely @xmath39 .
this latter constraint yields @xmath40 @xmath41 s@xmath36 .
a subtlety is that the transport would not be a strictly isotropic diffusion but could be the result of down pumping by anisotropic convection @xcite .
one way to infer the turbulent diffusivity in evolved stars is to note that isotopic anomalies in low - mass red giant branch and agb stars require that material from the base of the convective zone be transported to near the h - burning shell , processed , and returned to the convective envelope @xcite .
this so - called `` cool bottom processing '' ( cbp ) is thought to be magnetically driven @xcite . in the magnetic mixing scenario , a lower bound on the turbulent diffusivity is @xmath0@xmath42 @xmath41 s@xmath36 @xcite . while the actual diffusion coefficients in evolved star interiors are unknown , the values inferred from cbp are more than an order of magnitude lower than what are required for the envelope dynamo scenario to transport @xmath010@xmath43-g fields all the way to the wd surface .
higher field strengths ( in particular , @xmath0 mg fields ) lead to shorter buoyant - rise times and smaller cycle half - periods , and therefore require even greater diffusivity . in short , a substantial macroscopic diffusivity
would be needed to transport megagauss fields to the wd surface in agb stars .
we presently do not have independent constraints on these coefficients .
if the field generated in the envelope can not diffuse to the wd surface , an alternative possibility is amplification in an accretion disk that forms when the companion is tidally disrupted inside the common envelope @xcite . as shown below
, this scenario is attractive because it provides a natural mechanism for transporting the field to the proto - wd surface .
we consider disks formed from companions spanning the range from sub - jupiter - mass planets ( see below ) to brown dwarfs to low - mass stars ( i.e. @xmath44 between @xmath0@xmath45 and a few times @xmath46 ) . is the mass of jupiter .
this classification as planet or brown dwarf is based solely on mass @xcite ] tidal disruption of the companion results in formation of a disk inside the agb star .
this occurs at the tidal shredding radius , which we estimate as @xmath47 , where @xmath48 and @xmath44 are the radius and mass of the companion and @xmath49 is the stellar mass interior to @xmath50 @xcite .
we note that , since planetary and low - mass stellar companions to solar - type stars are significantly more plentiful than brown dwarfs @xcite , disks at the upper or lower mass range may be more common than those in the intermediate mass range .
the disk is ionized and susceptible to the development of magnetized turbulence ( e.g. via the magneto - rotational instability @xcite ) .
accretion towards the central proto - wd occurs on a viscous timescale given by @xmath51 , where @xmath52 is the keplerian orbital period at radius @xmath53 , @xmath54 is the effective kinematic viscosity , @xmath55 is the disk scaleheight , @xmath56 is the midplane soundspeed , @xmath57 is the keplerian orbital frequency , and @xmath58 is the dimensionsless shakura - sunyaev parameter that characterizes the efficiency of angular momentum transport @xcite .
the initial accretion rate can be approximated as @xmath59 where we have scaled @xmath48 to the radius of jupiter is the radius of jupiter .
note that @xmath48 depends only weakly on mass for the companions we consider . ] and have scaled @xmath60 to a large value @xmath61 because the disk can not cool efficiently at such high accretion rates and would be geometrically thick . since the companions under consideration lead to disks much more dense than the stellar envelope
, it is reasonable to ignore any interaction between the disk and the star
. equation ( [ eq : mdot0 ] ) illustrates that for @xmath62 , and for typical values of @xmath63 , @xmath64 is @xmath03 to 9 orders of magnitude larger than the eddington accretion rate of the proto - wd ( i.e. , @xmath65 yr@xmath36 ) . at first
, it might seem apparent that inflow of mass onto the wd surface would be inhibited by radiation pressure in such a scenario . however , this neglects the fact that , at sufficiently high accretion rates , photons are trapped and advected to small radii faster than they can diffuse out @xcite . in this ` hyper - critical ' regime , accretion is possible even when @xmath66 .
we evaluate the possibility of hyper - critical accretion by estimating the radius interior to which the inward accretion timescale , @xmath67 , is less than the timescale for photons to diffuse out of the disk midplane , @xmath68 ( which is approximately @xmath69 , where @xmath70 is the vertical optical depth @xcite ) .
this `` trapping radius '' is given by @xmath71 where @xmath72 is the opacity .
an important quantity is the ratio of the trapping radius to the outer disk radius @xmath50 ( coincident with the tidal shredding radius ) : @xmath73 where we have used equation ( [ eq : mdot0 ] ) and scaled @xmath72 to the electron scattering opacity @xmath74 @xmath35 g@xmath36 .
if @xmath75 , then @xmath76 and photons are advected with the matter . from equation ( [ eq : trap_ratio ] )
we conclude that @xmath77 for @xmath78 .
this implies that photon pressure will be unable to halt accretion initially and that the local energy released by accretion must be removed by advection @xcite .
advection acts like a conveyor belt , nominally carrying the gas to small radii as its angular momentum is removed .
if , however , there is no sink for the hot gas , this conveyor may jam " .
this is an important distinction between white dwarfs and neutron stars or black holes , as the latter two can remove the thermal energy by neutrino cooling or advection through the event horizon , respectively .
in contrast to a wd , neutrino cooling is ineffective and thermal pressure builds , such that radiation pressure may again become dynamically significant @xcite .
this in turn shuts down the accretion to at most , the eddington rate .
( eq . [ [ eq::toroidal ] ] ) , at the wd surface as a function of the mass of the tidally disrupted companion @xmath44 .
toroidal field strengths are presented for two values of the viscosity , @xmath79 ( _ solid line _ ) and @xmath80 ( _ dotted line _ ) and assuming an accretion efficiency @xmath81 .
the white dwarf mass and radius are @xmath82 and @xmath83 cm , respectively .
the vertical lines show the companion mass above which photons are trapped in the accretion flow ( i.e. , @xmath84 ; eq . [ [ eq : trap_ratio ] ] ) , such that super - eddington accretion occurs .
[ fig : fig1],width=336 ] however , this scenario neglects the possibility of outflows , which can sustain inflow by carrying away the majority of the thermal energy , thereby allowing a smaller fraction of the material to accrete at a higher than eddington rate . though more work is needed to assess the efficiency of outflows in the present context , radiatively - inefficient accretion flows are widely thought to be prone to powerful outflows @xcite .
even if accretion occurs at , or near , the eddington rate , the fields produced ( @xmath85 mg ) are still strong enough to explain the bulk of the magnetized wd population .
the origin of the strongest field systems ( @xmath0100 - 1000 mg ) may be problematic if accretion is limited to the eddington rate .
nevertheless , the material deposited outside the wd will be hot and virialized , forming an extended envelope with a lengthscale comparable to the radius .
though initially hot , this material will eventually cool ( on longer , stellar timescales ) and become incorporated into the stellar layers near the proto - wd surface .
if this material cools at some fraction of the local eddington luminosity , it will be incorporated onto the wd surface on a timescale @xmath0@xmath86 years , which is much less than typical agb lifetimes .
a final complication may arise because the composition of the companion is likely to be hydrogen - rich and will begin to burn thermonuclearly as material accretes or after it settles on the proto - wd surface . because the energy released in burning hydrogen to helium ( @xmath87 mev nucleon@xmath36 ) vastly exceeds the gravitational binding energy per nucleon at the wd surface ( @xmath88 mev nucleon@xmath36 ) , burning that occurs explosively could have a dynamical impact . for the accretion rates of relevance
, we estimate that the midplane temperature of the disk exceeds @xmath89 k , such that hydrogen will burn via the hot cno cycle .
the hot cno cycle occurs at a rate that is independent of temperature and is itself thermally stable . however , hydrogen - rich material deposited in the he burning layer may , under some circumstances , trigger a thermonuclear runaway @xcite , which could remove the accreted mass from the proto - wd surface , or even eject the common envelope entirely . in what follows
we assume stable burning and set aside this important caveat . additional work will , however , be required to address what conditions are required for stable versus explosive burning . due to the presence of shear in the disk ,
the mri is a likely source of turbulence , which in turn amplifies magnetic field .
large - scale fields produced by the mri have been modeled via @xmath5 dynamos at various levels of sophistication ( see @xcite for the most recent example ) . however , for the purposes of estimating orders of magnitude of the fields , approximate values from less sophisticated treatments can be employed .
although this situation is qualitatively similar to that described in the previous subsection , the dynamo we envision here operates in the accretion disk and not the envelope .
the alfvn velocity in the disk obeys @xmath90 @xcite , such that the mean toroidal field at radius @xmath53 is given by @xcite @xmath91 where in the second equality we have substituted equation ( [ eq : mdot0 ] ) for @xmath64 , and multiplied by the factor @xmath92 to account for the possibility of outflows as described above . if an @xmath5 dynamo operates , the mean poloidal field @xmath93 is related to the toroidal field via @xmath94 @xcite .
however , regardless of whether a large scale field is generated , a turbulent field of magnitude @xmath95 is likely to be present and contributing significantly to the maxwell stresses responsible for disk accretion . figure [ fig : fig1 ] shows the toroidal field evaluated near the wd surface @xmath96 cm as a function of companion mass @xmath44 , calculated for two different values of the viscosity ( @xmath79 and @xmath97 ) . in both cases ,
we assume that @xmath81 .
note that for the range of relevant companion masses @xmath98 @xmath99 , @xmath100 mg , in precisely the correct range to explain the inferred surface field strengths of hfmwds .
although our results suggest that companion accretion can produce the field strengths necessary to explain hfmwds , for the field to be present on the surface of the wd when it forms , it must at a minimum survive the remaining lifetime of the star .
in particular , the field will decay due to ohmic diffusion on the timescale @xmath101 where @xmath102 @xmath35s@xmath36 is the spitzer resistivity , ln@xmath103 is the coulomb logarithm , @xmath104 is the stellar temperature near the proto - wd surface , and @xmath105 is the final thickness of the accreted companion mass after it is incorporated into the surface layers of proto - wd , where @xmath106 is the mean density of the proto - wd . equation ( [ eq : taudecay ] ) shows that for typical core temperatures during the agb phase @xmath107 k , @xmath108 exceeds the mean agb lifetime @xmath109 myr for @xmath110 .
this suggests that fields may survive ohmic decay , at least for very massive companions .
furthermore , the magnetized companion material may become incorporated into the degenerate wd core long before the agb phase ends , in which case the much higher conductivity due to degenerate electrons substantially increase the ohmic decay timescale over that given by equation ( [ eq : taudecay ] ) , thereby ensuring long - term field survival .
recent observational evidence from the sloan digital sky survey strongly suggests that high - field magnetic white dwarfs originate from binary interactions .
in particular , it was proposed that the progenitors of hfmwds are binary systems that evolve through a common envelope phase @xcite . in this scenario , companions that survive ce evolution ( leaving tight binaries ) may produce magnetic cataclysmic variables , while those that do not survive ( i.e. that merge ) may produce isolated hfmwds . to investigate this hypothesis ,
we have estimated the magnetic fields resulting from a low - mass companion embedded in a ce phase with a post - ms giant . during in - spiral , the companion transfers energy and angular momentum to the envelope .
the resulting shearing profile , coupled with the presence of convection , can amplify large - scale magnetic fields via dynamo action in the envelope .
we incorporate the back - reaction of the magnetic field growth on the shear , which results in a transient dynamo .
the fields generated at the interface between the convective and radiative zones are weak ( @xmath02@xmath111 g ) and would have to diffuse to the wd surface and anchor there if they are to explain the hfmwds .
a successful envelope dynamo scenario must also survive to the end of the agb phase , which probably means only if the companion ejects the turbulent envelope ; the dynamo is transient as long as there is a source of turbulent diffusion .
the bulk of the systems formed this way might be post - ce , tight binaries potentially magnetic cvs .
to explain isolated hfmwds , an alternative to a dynamo operating in the envelope is a dynamo operating in an accretion disk . during the ce phase ,
if the companion is of sufficiently low mass , it avoids prematurely ejecting the envelope .
instead , in - spiral proceeds until the companion is tidally shredded by the gravitational field of the proto - wd .
the subsequent formation of an accretion disk , which also posesses turbulence and shear , can amplify magnetic fields via dynamo action . for the range of disrupted companions considered here , the disk initially accretes at super - eddington values . in this hyper - critical regime ,
the timescale for photon diffusion out of the disk is longer than the viscous timescale . for a range of disrupted - companion disks , we find that the saturated toroidal mean - field attains values between a few and a few thousand megagauss .
amplification of the magnetic field in a super - eddington accretion disk is attractive as it reproduces the observed range of hfmwds and naturally transports magnetic flux to the wd surface .
high - mass stars may also undergo common envelope interactions in the presence of close companions .
if the common envelope field mechanisms described here operate in high - mass stars , then the result could be strong field neutron stars or magnetars ( neutron stars with magnetic fields in excess of @xmath0@xmath112 g ) .
in particular , formation of an accretion disk from an engulfed companion during a red supergiant phase could produce a magnetized wd core . in the eventual core - collapse and stellar supernova explosion ( possibly driven by the neutrino mechanism ;
@xcite ) , the magnetized wd core collapses to a neutron star . if simple flux freezing operates ( itself an open question ) and the initial magnetized core is on the order of @xmath0100 - 1000 mg , homologous collapse to a neutron star would generate @xmath0@xmath112 g fields .
typical neutron stars that possess modest field strengths may originate from core - collapse supernova of single stars or stars without having incurred a ce phase .
note that what we are proposing here is an alternative to the neutrino - driven convection dynamo described in @xcite . in our model ,
the engulfment of a companion and the formation of an accretion disk naturally provides fast rotation , magnetized turbulence and differential rotation .
we emphasize that the viability of this mechanism depends on the length scale of the magnetic field deposited in the pre - collapse core , which must be sufficiently large to produce , upon collapse , the dipole - scale , volume - encompassing fields necessary to account for the spin - down behavior of magnetars and the energy budget of their giant flares @xcite . in summary , common envelope evolution as the origin of strongly - magnetized
, compact objects seems plausible . whether this hypothesis is ultimately found to be viable
will depend on the statistics of low - mass stellar and substellar companions to stars of similar masses to ( or somewhat higher masses than ) the sun .
the numerous radial velocity searches of the last 20 years have revealed a number of such companions @xcite .
the precise occurrence rate of companions in orbits that could lead to the kind of disk - dynamo mechanism described above remains unclear ( though if such companions turn out to be rarer than the hfmwd fraction , then the sdss data indicating a binary origin would be puzzling ) . further theoretical work into the binary origin of hfmwds and magnetars requires the development of multi - dimensional , magnetohydrodynamic simulations of the ce phase .
such an approach has already had initial success for purely hydrodynamic adaptive mesh refinement simulations @xcite .
we thank alberto lopez , jay farihi , jim stone , jeremy goodman , kristen menou , deepak raghavan , andrei mesinger , adam burrows and fergal mullally for thoughtful discussions .
jn acknowledges support for this work from nasa grant hst - ar-12146 .
dss acknowledges support from nasa grant nnx07ag80 g .
support for bdm is provided by nasa through einstein postdoctoral fellowship grant number pf9 - 00065 awarded by the chandra x - ray center , which is operated by the smithsonian astrophysical observatory for nasa under contract nas8 - 03060 .
egb acknowledges support from grants nsf phy-0903797 and nsf ast-0807363 . | the origin of highly - magnetized white dwarfs has remained a mystery since their initial discovery .
recent observations indicate that the formation of high - field magnetic white dwarfs is intimately related to strong binary interactions during post - main - sequence phases of stellar evolution . if a low - mass companion , such as a planet , brown dwarf , or low - mass star is engulfed by a post - main - sequence giant , gravitational torques in the envelope of the giant lead to a reduction of the companion s orbit .
sufficiently low - mass companions in - spiral until they are shredded by the strong gravitational tides near the white dwarf core .
subsequent formation of a super - eddington accretion disk from the disrupted companion inside a common envelope can dramatically amplify magnetic fields via a dynamo . here
, we show that these disk - generated fields are sufficiently strong to explain the observed range of magnetic field strengths for isolated , high - field magnetic white dwarfs .
a higher - mass binary analogue may also contribute to the origin of magnetar fields . |
the study of the electromagnetic structure of the baryons proves to be an important source for understanding their inner structure , as well as for obtaining information on strong interaction in the confinement region .
the electromagnetic structure of baryons is described by the form factors whose experimental and theoretical investigations have been continuing for a long time up today .
operation of new electron beam facilities at jefferson laboratory ( jlab ) , mainz microton ( mami ) at mainz , and mit / bates open new horizons in study of the form factors of nucleons and their excitations .
the nucleon excitations are studied in reactions such as scattering of electron beams or target photons from nucleons , i.e. , @xmath6 , where @xmath7 is the nucleon excitation and @xmath8 is the virtual photon .
obviously , more information can be obtained about the structure of baryons at small distances with the increasing photon virtuality .
it is hoped that the new electron beam facilities would definitely allow collecting large amount of more precise and complete data in the studies of the electro - excitations of nucleon resonances . along these lines essential information
has already been obtained , in particular , for the spin-3/2 and negative parity @xmath9 state in the range up to @xmath10 .
progress in the experiments has simulated new significant developments in understanding the dynamics of qcd at large distance from theoretical side ( for more detail see @xcite ) . in the present work we study the @xmath11 transition form factors in framework of the light cone qcd sum rules ( lcsr ) @xcite .
this transition was studied within the framework of the nonrelativistic and relativistic quark model ( see @xcite and references therein ) , single quark transition model @xcite , and the covariant spectator quark model @xcite , respectively .
the experimental results for the transition form factors were obtained from clas data @xcite .
recently new data has been obtained in a wider @xmath4 domain , which surely has higher accuracy compared to the previous experimental results .
the paper is organized as follows : in section 2 the sum rules for the form factors responsible for the @xmath11 transition are obtained within the lcsr method . in this section
we also present the expressions for the helicity amplitudes which describe this transition . in section 3 numerical results for the transition form factors , as well as for the helicity amplitudes
are presented .
this section contains also discussions and conclusions .
in order to derive the light cone qcd sum rules for the @xmath12 we introduce the following correlation function : [ ebozo01 ] _ ( p , q ) = i d^4x e^iqx 0 t \{_(0 ) j_^el(x ) } n(p ) , where @xmath13 is the interpolating current of the spin-3/2 positive parity baryons , and @xmath14 is the electromagnetic current defines as @xmath15 .
the interpolating current of the spin-3/2 baryons is given by the following expression : [ ebozo02 ] _
= ^abc \ { ( q_1^at c _ q_2^b ) ^ _ q_3^c - ( q_1^at c _ q_3^b ) ^ _ q_2^c } , where @xmath16 , @xmath17 for the excited proton , and @xmath18 , @xmath19 for the excited neutron states , respectively ; and @xmath20 , @xmath21 , and @xmath22 are the color indices and @xmath23 is the charge conjugation operator .
we start our discussion by considering the hadronic transition involving @xmath9 . within the framework of lcsr method
this is implemented by calculating the correlation function from the hadronic and quark - gluon sides , respectively , and then equating these representations . in calculating the correlation function from the hadronic side , the total set of hadrons which carry the same quantum numbers of the interpolating current
should be inserted in it .
the contributions of the lowest spin-3/2 negative and positive parity states can be obtained by saturating the hadronic part of the correlation function with them , from which we get : [ ebozo03 ] _ 0 _ ( p^ ) ( p^ ) j_^el n(p ) , + 0 _ ( p^ ) ( p^ ) j_^el n(p ) , + 0 _ ( p^ ) ( p^ ) j_^el n(p ) , + 0 _ ( p^ ) ( p^ ) j_^el n(p ) .
it should be noted here that the current @xmath13 interacts not only with spin-3/2 states , but also spin-1/2 states , and therefore we consider their contributions on eq .
( [ ebozo03 ] ) .
we start by considering the last two matrix elements which correspond to the contributions of the spin1/2 states . in the general case ,
the matrix element @xmath24 is determined as follows : [ ebozo04 ] 0 _ ^+ ( p^ ) = ( a p_^+ b _ )
u(p^ ) . multiplying both sides of eq .
( [ ebozo04 ] ) and using the fact that @xmath25 , we immediately obtain that : [ ebozo05 ] 0 _ ^+ ( p^ ) = b ( _ - 4 m _
p_^)u(p^ ) . similarly , from the parity consideration the matrix element @xmath26 is determined as : [ ebozo06 ] 0 _ ^- ( p^ ) = b _ 5 ( _ - 4 m _
p_^ ) u(p^ ) .
it follows from eqs .
( [ ebozo05 ] ) and ( [ ebozo06 ] ) that the contributions of the spin-1/2 states are proportional to either @xmath27 or @xmath28 , which we will use later .
the remaining matrix elements are determined as follows : [ ebozo07 ] ( p^ ) j_^el ( p ) = |u(p^ ) u(p ) , + [ ebozo08 ] ( p^ ) j_^el ( p ) = |u(p^ )
_ 5 u(p ) , where @xmath29 state corresponds to the one nucleon state .
the matrix elements which describe the contributions of the spin-3/2 states are defined as : [ ebozo09 ] 0 j _ ( p^ ) _
+ u_(p^ ) , + 0 j _ ( p^ ) _ - _ 5 u_(p^ ) , + [ ebozo10 ] ( p^ ) j_^el ( p ) |u_(p^ ) \ { g_1^+ ( q__- /q g _ ) + g_2^+ [ q__- ( q p )
g _ ] + g_3^+ ( q_q_- q^2 g _ ) } _ 5 u(p ) , + [ ebozo11 ] ( p^ ) j_^el ( p ) |u_(p^ ) \ { g_1 ^ - ( q__- /q g _ ) + g_2 ^ - [ q__- ( q p ) g _ ] + g_3 ^ - ( q_q_- q^2 g _ ) } u(p ) , where @xmath30 ; @xmath31 ; and @xmath32 are the form form factors . using the definitions of the residues and form factors given in eqs .
( [ ebozo09])-([ebozo11 ] ) , and summing over spins of spin-3/2 states given as : [ nolabel01 ] u_(p^ ) |u_(p^ ) ( /p^+ m _ ) , for the phenomenological part we obtain , [ ebozo12 ] _ - _ - m_-p^2 _ 5 ( /p^+ m _ ) + \ { g_1 ^ - ( q__- /q g _ ) + g_2 ^ - [ q__- ( q p ) g _ ] + g_3 ^ - ( q_q_- q^2 g _ ) } u(p ) + ( /p^+ m _ ) + \ { g_1^+ ( q__- /q g _ ) + g_2^+ [ q__- ( q p ) g _ ] + g_3^+ ( q_q_- q^2 g _ ) } _ 5 u(p ) + ( _ - 4 m _
p_^ ) ( /p^+ m _ ) u(p ) + _ 5 ( _ - 4 m _
p_^ ) ( /p^+ m _ ) _ 5
u(p ) one can easily see from eq .
( [ ebozo12 ] ) that , in the first two terms which describe the contributions of the spin-3/2 states , all structures , except @xmath33 contain also contributions coming from spin-1/2 states which should all be removed .
we see from eq .
( [ ebozo12 ] ) that the terms with @xmath28 at left or proportional to @xmath27 , as well as @xmath34 at right or terms proportional to @xmath35 , describe the contributions of the spin-1/2 states .
consider the terms multiplied by the structure @xmath36 which can be written as @xmath37 .
both terms on the right in this identity contain the contributions coming from spin-1/2 states . as a result
one can conclude that , in eq .
( [ ebozo12 ] ) only the terms that are proportional to the @xmath33 structure describe the contributions of spin-3/2 states .
so , the hadronic part of the correlation function that involves contributions from both negative and positive parity spin-3/2 states only can be written as : [ ebozo13 ] _ ( /p^+ m _ ) _ 5 u(p ) + ( /p^+ m _ ) _ 5 u(p ) .
we see from this equation that the correlation function contains numerous lorentz structures all of which are not independent of each other . in order to allocate independent structures ordering of the dirac matrices
is implemented , from which we choose the ordering @xmath38 .
the correlation function ( [ ebozo01 ] ) can be decomposed in terms of the independent invariant amplitudes as follows : [ ebozo14 ] _ _ 1 [ ( p+q)^2,q^2 ] /q _ _ 5 q_+ _ 2 [ ( p+q)^2,q^2 ] _ _ 5 q_+ _ 3 [ ( p+q)^2,q^2 ] /q _ 5 p_q _ + _ 4 [ ( p+q)^2,q^2 ] _ 5 p_q_+ _ 5 [ ( p+q)^2,q^2 ] /q _ 5 q_q_+ _ 6 [ ( p+q)^2,q^2 ] _ 5 q_q _ + comparing eqs .
( [ ebozo13 ] ) and ( [ ebozo14 ] ) , we get the following set of equations from which we can determine the form factors @xmath39 .
[ ebozo15 ] - _ - m_-p^2 g_1 ^ - - _ + m_-p^2 g_1^+ _ 1 + _ - m _
m_-p^2 g_1 ^ - - _ + m_m_-p^2 g_1^+ _ 2 + _ - m_-p^2 g_2 ^ - - _ + m_-p^2 g_2^+ _ 3 + - _ - m _
m_-p^2 g_2 ^ - - _ + m_m_-p^2 g_2^+ _ 4 + _ - m_-p^2 ( g_2 ^ -2 - g_3 ^ - ) - _ + m_-p^2 ( g_2 ^ - 2 - g_3^+ ) _ 5 + - _ - m _ m_-p^2 ( g_2 ^ -2 - g_3 ^ - ) - _ + m _ m_-p^2 ( g_2^+ 2 - g_3^+ ) _ 6 . from these equations
we finally obtain the following expressions for the transition form factors @xmath40 : [ ebozo16 ] _ - ( m_+ m _ ) m_-p^2 g_1 ^ - - m _ _ 1 + _ 2 + _ - ( m_+ m _ ) m_-p^2 g_2 ^ - m _ _ 3 - _ 4 + _ - ( m_+ m _ ) m_-p^2 ( g_2 ^ - 2 - g_3 ^ - ) m _ _ 5 - _ 6 . using the quark hadron duality ansatz
, we next calculate the invariant functions @xmath41 from the qcd side . in order to justify the operator product expansion ( ope ) near the light cone @xmath42
, the external momenta @xmath43 and @xmath44 are taken to be space - like .
the result of ope is obtained in terms of the nucleon distribution amplitudes ( das ) of growing twist .
the nucleon das are the key ingredient of the lcsr and they are calculated in @xcite up to twist-6 .
having the explicit forms of the nucleon das , the the invariant amplitudes @xmath45 can be written in the following form : [ ebozo17 ] _
i[(p+q)^2,q^2 ] = _
i=1 ^ 3 _ 0 ^ 1 _ i , n
[ x , q^2,(p+q)^2 ] ^n , where @xmath46 ; @xmath47 .
the expressions of @xmath48 for the form factors @xmath49 , @xmath50 and @xmath51 are presented in appendix a. substituting the results for @xmath41 given in eq .
( [ ebozo17 ] ) into eq .
( [ ebozo16 ] ) , and performing borel transformation with respect to @xmath43 in order to suppress contributions of the higher states and continuum , we obtain the following sum rules for the form factors @xmath52 , @xmath53 and @xmath54 : [ ebozo18 ] _ - ( m _ + m _ ) g_1 ^ -(q^2 ) e^-m_/m^2 - m _
i_1(q^2,m^2,s_0 ) + i_2(q^2,m^2,s_0 ) , + _ - ( m _ + m _ ) g_2 ^ -(q^2 ) e^-m_/m^2 m _ i_3(q^2,m^2,s_0 ) - i_4(q^2,m^2,s_0 ) , + _ - ( m _ + m _ )
e^-m_/m^2 m _
i_5(q^2,m^2,s_0 ) + i_6(q^2,m^2,s_0 ) , where [ ebozo19 ] i_i(q^2,m^2,s_0 ) _
x_0 ^ 1 dx e^-s(x)/m^2 + e^-s_0/ m^2 } , where @xmath55 is the borel mass square , @xmath56 is the continuum threshold and [ nolabel02 ] s(x ) = |x q^2 + x |x m_b^2 x , @xmath57 , and @xmath58 is the solution of the equation @xmath59 .
it follows from eq .
( [ ebozo19 ] ) that the residue of the negative parity spin-3/2 @xmath9 baryon is needed .
the decay constant @xmath60 of the negative parity @xmath9 baryon can be obtained from the following two - point correlation function [ ebozo20 ] _ ( p^2 )
i d^4x e^ipx 0 _ ( x ) _ 1(p^2 ) g _ + _ 2(p^2 ) p_p_+ _ 3(p^2 ) _ _ + _ 4(p^2 ) _ p_+ _ 5(p^2 ) _ p _ .
as we have already noted , the structures @xmath61 , @xmath62 , @xmath63 and @xmath64 contain contributions from spin-1/2 states .
therefore , in order to determine the decay constant of the negative parity spin-3/2 baryon , the above - mentioned structures should be omitted .
the only structure which contains contributions from spin-3/2 states with both parity is @xmath65 . retaining only this structure , and saturating ( [ ebozo20 ] ) with spin-3/2 states , from the physical part of the correlation function
we get , [ ebozo21 ] _ + ^2 ( p + m _ ) m_-p^2 + _ -^2 ( p - m _ ) m_-p^2 + denoting the spectral densities @xmath66 and @xmath67 corresponding to the structures @xmath68 and @xmath33 , respectively , and performing borel transformation over @xmath69 we get , [ ebozo22 ] _ + ^2 e^-m_/m^2 + _
-^2 e^-m_/m^2 _ 0^s_0 ds e^-s / m^2 _ 1(s ) , + m _ _ + ^2 e^-m_/m^2 - _ -^2 m _ e^-m_/m^2 _ 0^s_0 ds e^-s / m^2 _ 2(s ) . solving these equations for the mass and decay constant of the negative parity spin-3/2 @xmath70 baryon , we get [ ebozo23 ] m _ + [ ebozo24 ] _
-^2 _ 0^s_0 ds e^-s / m^2 . in order to calculate the the spectral densities @xmath71 and @xmath72 , we need to obtain the expressions @xmath73 and @xmath74 corresponding to the structures @xmath68 and @xmath33 , respectively .
the explicit forms of @xmath73 and @xmath74 after the borel transformation is performed , are given in appendix b ( see also @xcite ) .
the spectral densities @xmath66 and @xmath67 can be obtained after applying the fourier transformation and performing continuum subtraction to these expressions .
in general , the electromagnetic properties of hadrons can also be described in terms of the helicity amplitudes or multipole electromagnetic form factors instead of the form factors @xmath75 , @xmath76 and @xmath77 .
the multipole electromagnetic form factors in our case are the magnetic dipole @xmath78 , electric quadrupole @xmath79 and coulomb quadrupole @xmath80 .
these form factors are defined as ( see for example @xcite ) : [ ebozo25 ] g_m -1 m_nm _ - m_n , + [ ebozo26 ] g_e -1 m_nm _ - m_n + [ ebozo27 ] g_c -1 m_nm _ - m_n the relations among helicity amplitudes and multipole form factors are given as follows : [ ebozo28 ] a_1/2 ( 3 g_m - g_e ) + [ ebozo29 ] a_3/2 ( g_m + g_e ) + [ ebozo30 ] s_1/2 - e 4 m _
g_c , where [ nolabel03 ] c , + + g_c 4 m _
g_1 + ( 3 m_^2 + m_n^2 + q^2 ) g_2 + 2 ( m_^2 - m_n^2 - q^2 ) g_3 , + + is the magnitude of the nucleon three - momentum in the rest frame of the @xmath70 baryon .
here in this section we present the numerical results of the lcsr prediction on the multipole form factors , as well as helicity amplitudes for the @xmath0 transition .
the main input parameters of this method are the distribution amplitudes ( das ) , and we use the results of @xcite in the following numerical calculations .
the values of the factors appearing in the nucleon das are calculated within the lcsr , and they are given as : [ nolabel05 ] f_n ( 5.0 0.5 ) 10 ^ -3 gev^2 , + _ 1 - ( 2.7 0.9 ) 10 ^ -2 gev^2 , + _ 2 - ( 5.4 1.9 ) 10 ^ -2 gev^2 , and @xmath81 , @xmath82 , @xmath83 , and @xmath84 @xcite .
in calculation of the transition form factors responsible for the @xmath0 transition masses and decay constants of the negative parity baryons , which are determined from the analysis of the two - point sum rules ( see eq .
( [ ebozo22 ] ) ) .
these sum rules have two auxiliary parameters , namely , the borel mass parameter @xmath55 and the continuum threshold @xmath56 .
the working window for the borel mass parameter is determined from the condition that the nonperturbative and continuum contribution should be less than , say , half of the perturbative contributions .
the value of the continuum threshold is determined from the condition that , the mass sum rules reproduces the experimental value of the @xmath9 mass with @xmath85 accuracy .
the analysis of the mass sum rules shows that the criteria for the borel mass parameter which has been noted above , is fulfilled in the region @xmath86 .
we also observe that if the continuum threshold is chosen in the region @xmath87 , sum rules reproduce the mass of the @xmath9 baryon with @xmath85 accuracy , as the result of which we get @xmath88 .
in further analysis we shall use this value of @xmath89 , and the above - mentioned regions of @xmath55 and @xmath56 . in figs . 1 , 2 and 3 we present the dependencies of the helicity amplitudes @xmath90 , @xmath91 and @xmath92 on @xmath4 , respectively , at several fixed values of the borel parameter @xmath55 and at @xmath93 . in the numerical analysis
we take into account the errors in the input parameters and the graphs are presented at the minimum values of the input parameters . in order to keep the higher twist and higher states contributions under control , @xmath4 is running in the region @xmath94 .
from all these we figures we see that , the trend of behavior for all three figures are the same , i.e. , all three amplitudes tend to zero with increasing @xmath4 .
moreover , @xmath91 is always positive , while @xmath90 and @xmath92 are alway negative for all values of @xmath4 .
if the maximum values of the input parameters are taken into account , the values of the helicity amplitudes increase slightly in modulo . among all these three amplitudes , @xmath90 is less sensitive to the errors in the input parameters .
when we compare our results with the predictions of the spectator quark model @xcite , we observe that our results on the value of @xmath90 are approximately the same as with the prediction of @xcite , and very close to the experimental result @xcite . in the case of @xmath92 amplitude ,
our results are slightly smaller compared those given in @xcite .
this result demonstrates the necessity for more precise determination of the input parameters . in the case for the @xmath91 amplitude ,
our results are very close to the ones given in @xcite within the limits of errors , as well as to the experimental data given in @xcite .
our prediction on the form factors and helicity amplitudes can be refined further by taking the radiative corrections into account for the distribution amplitudes .
the first calculation along this line has been carried out in @xcite . in summary , we study the @xmath0 transition within the lcsr method .
we derive the sum rules for the magnetic dipole @xmath78 , electric quadrupole @xmath79 , and coulomb quadrupole @xmath80 form factors , and consequently for the helicity amplitudes @xmath90 , @xmath91 and @xmath92 .
the @xmath4 dependence of these helicity amplitudes are investigated .
finally , we compare our results with the predictions of the covariant spectator quark model , and with the existing experimental data .
in this appendix we present the expressions for the functions @xmath72 , @xmath95 and @xmath96 which appear in the sum rules for @xmath49 , @xmath50 , and @xmath97 .
_ 6 ( x)-32 ( 1-x ) m_n^2 m _ ( x^2 m_n^2 + q^2 ) ( e_q_1 _ 6 + e_q_2 _ 6 ) , + + _ 4 ( x ) 8 e_q_1 m_n^2 + 8 e_q_2 m_n^2 + 16 e_q_3 m_n^2 + 8 e_q_1 1-xx m_n \ { ( x^2 m_n^2 + q^2 ) _ 2 + _ 4 + 2 x } + 8 e_q_2 1-xx m_n \ { ( x^2 m_n^2 + q^2 ) _ 2 + _ 4 + 2 x } + 8 e_q_3 ( 1-x ) m_n + 16 e_q_1 m_n^2 m _
_ 0^|x dx_3 ( x,1-x - x_3,x_3 ) + 16 e_q_2 m_n^2 m _ _ 0^|x dx_1 ( x_1,x,1-x_1-x ) + 16 e_q_3 m_n^2 m _ _ 0^|xdx_1 ( x_1,1-x_1-x , x ) + + _ 2 ( x ) - 8 e_q_1 m_nx + 8 e_q_2 m_nx + 16 e_q_1 _ 0^|x dx_3 \ { ( 1-x ) m_n ( a_3-p_1+s_1+v_3 ) + m _ } ( x,1-x - x_3,x_3 ) + + 16 e_q_2 _ 0^|x dx_1 \ { ( 1-x ) m_n ( p_1-s_1 ) + m _ } ( x_1,x,1-x_1-x ) + + 16 e_q_3 _ 0^|xdx_1 \ { ( 1-x ) m_n ( a_3+v_3 ) + m _ } ( x_1,1-x_1-x , x ) .
_ 6(x ) -64 e_q_1 ( 1-x ) m_n^2 \ { _ 6 + 2 x^2 m_n m _ ( 2 _ 8 + _ 6 + _ 6 ) } + 64 e_q_2 ( 1-x ) m_n^2 \ { _ 6 + 2 x^2 m_n m _ ( 2 _ 8 + _ 6 + _ 6 ) } + 128 e_q_3 x ( 1-x ) m_n^2 \ { ( x m_n^2 + q^2 ) _ 6 + x m_n m _ ( 2 _ 8 + _ 6 + _ 6 ) } + + _ 4 ( x ) 16 e_q_1 ( 3 - 4 x ) m_n^2 _ 6 + 16 e_q_2 ( 3 - 4 x ) m_n^2 _ 6 + 16 e_q_1 x m_n \ { ( 1-x ) m_n + 2 m _ } + 16 e_q_2 x m_n \ { ( 1-x ) m_n + 2 m _ } + 32 e_q_3 x m_n \ { ( 1-x ) m_n ( 2 _ 5 + _ 5 - _ 5 ) - m _ } + 32 e_q_1 m_n^2 _ 0^|x dx_3 ( 1-x ) ( x,1-x - x_3,x_3 ) + 32 e_q_2 m_n^2 _ 0^|x dx_1 ( x_1,x,1-x_1-x ) + 64 e_q_3 m_n^2 _ 0^|xdx_1 x t_1^m ( x_1,1-x_1-x , x ) + + _ 2 ( x ) - 32 e_q_1 _ 0^|x dx_3 ( 1-x ) ( x,1-x - x_3,x_3 ) + 32 e_q_2 _ 0^|x dx_1 ( x_1,x,1-x_1-x ) + 64 e_q_3 _ 0^|xdx_1 x t_1 ( x_1,1-x_1-x , x ) .
_ 6 ( x ) - 128 e_q_1 ( 1-x)^2 m_n^2 + 128 e_q_2 ( 1-x)^2 m_n^2 + 128 e_q_3 ( 1-x)^2 m_n^2 + + _ 4 ( x ) - 32 e_q_1 ( 1-x ) m_n^2 _ 6 - 32 e_q_2 ( 1-x ) m_n^2 _ 6 - 32 e_q_3 ( 1-x ) m_n^2 _ 6 + 16 e_q_1 ( 1-x ) m_n \ { ( 1-x ) m_n ( 4 _ 5 + _ 4 + _ 5 - _ 4 - _ 5 ) + 2 m _ } + 16 e_q_2 ( 1-x ) m_n \ { ( 1-x ) m_n ( 4 _ 5 + _ 4 + _ 5 - _ 4 - _ 5 ) + 2 m _ } + 16 e_q_3 ( 1-x ) m_n \ { ( 1-x ) m_n ( 4 _ 5 + _ 4 + _ 5 - _ 4 - _ 5 ) + 2 m _ } + 32 e_q_1 m_n^2 _ 0^|x dx_3 ( 1-x ) ( x,1-x - x_3,x_3 ) + 32 e_q_2 m_n^2 _ 0^|x dx_1 ( 1-x ) ( x_1,x,1-x_1-x ) + 32 e_q_3 m_n^2 _ 0^|xdx_1 ( 1-x ) ( x_1,1-x_1-x , x ) + + _ 2 ( x ) 32 e_q_1 _ 0^|x dx_3 ( 1-x ) ( x,1-x - x_3,x_3 ) + 32 e_q_2 _ 0^|x dx_1 ( 1-x ) ( x_1,x,1-x_1-x ) + 32 e_q_3 _ 0^|xdx_1 ( 1-x ) ( x_1,1-x_1-x , x ) . in the above expressions for @xmath99 the functions @xmath100 are defined in the following way : [ nolabel06 ] ( x_1 ) _
1^x_1dx_1^_0 ^ 1- x^_1dx_3 f(x_1^,1-x_1^-x_3,x_3 ) , + ( x_1 ) _
1^x_1dx_1^_1^x^_1dx_1^ _ 0 ^ 1- x^_1dx_3 f(x_1^,1-x_1^-x_3,x_3 ) , + ( x_2 ) _
1^x_2dx_2^_0 ^ 1- x^_2dx_1 f(x_1,x_2^,1-x_1-x_2^ ) , + ( x_2 ) _ 1^x_2dx_2^_1^x^_2dx_2^ _ 0 ^ 1- x^_2dx_1 f(x_1,x_2^,1-x_1-x_2^ ) , + ( x_3 ) _
1^x_3dx_3^_0 ^ 1- x^_3dx_1 f(x_1,1-x_1-x_3^,x_3^ ) , + ( x_3 ) _ 1^x_3dx_3^_1^x^_3dx_3^ _ 0 ^ 1- x^_3dx_1 f(x_1,1-x_1-x_3^,x_3^ ) .
definitions of the functions @xmath101 , @xmath102 , @xmath103 , @xmath104 and @xmath105 that appear in the expressions for @xmath106 are given as follows : [ nolabel07 ] b_2 t_1+t_2 - 2 t_3 , + b_4 t_1-t_2 - 2 t_7 , + b_5 - t_1+t_5 + 2 t_8 , + b_6 2 t_1 - 2 t_3 - 2 t_4 + 2 t_5 + 2 t_7 + 2 t_8 , + b_7 t_7-t_8 , + b_8 -t_1+t_2+t_5-t_6 + 2 t_7 + 2t_8 , + c_2 v_1-v_2-v_3 , + c_4 -2v_1+v_3+v_4 + 2v_5 , + c_5 v_4-v_3 , + c_6 -v_1+v_2+v_3+v_4+v_5-v_6 , + d_2 -a_1+a_2-a_3 , + d_4 -2a_1-a_3-a_4 + 2a_5 , + d_5 a_3-a_4 , + d_6 a_1-a_2+a_3+a_4-a_5+a_6 , + e_1 s_1-s_2 , + h_1 p_2-p_1 .
note that in the numerical analysis for the form factors the masses of the light quarks are all set to zero .
in this appendix we present the expressions for the functions @xmath74 and @xmath74 corresponding to the structures @xmath68 and @xmath33 , respectively , after the borel transformation and continuum subtraction are performed . _ 13 ^4 m^6 e_2(x ) - 5 24 ^4 m^2
e_0(x ) + 28 9 m^2 m_0 ^ 2 ^2 - 16 ^2 + + _ 2 + 4 ^2 m^4 e_1(x ) - 2 ^2 m_0 ^ 2 m^2 e_0(x ) + 1 9 ^2 m^2 m_0 ^ 2 + , where [ nolabel08 ] e_n(x)=1-e^-x_i=0^n , with @xmath107 , and @xmath56 is being the continuum threshold .
99 i. g. aznauryan , a. bashir , v. braun , s. j. brodsky _ et .
e * 22 * , 1330015 ( 2013 ) .
v. m. braun , arxiv:9801222 [ hep - ph ] ( 1998 ) .
e. santopinto , and m. m. giannini , phys .
c * 86 * , 065202 ( 2012 ) .
b. golli , and s. sirca , eur .
j. a * 49 * , 111 ( 2013 ) . v. d. burkert , t. s. h. lee , int .
e * 13 * , 1035 ( 2004 ) .
g. ramalho , and m. t. pena , phys .
d * 89 * , 094016 ( 2014 ) .
i. g. aznauryan _ et .
al _ , clas collaboration , phys .
c * 80 * , 055203 ( 2009 ) . v. i. mokeev _ et .
al _ , clas collaboration , phys .
c * 86 * , 035203 ( 2012 ) .
v. m. braun , r. j. fries , n. mahnke , and e. stern , nucl . phys .
b * 589 * , 381 ( 2000 ) ; nucl .
b * 607 * , 433e ( 2001 ) .
frank x. lee , and xinyu liu phys .
d * 66 * , 014014 ( 2002 ) .
v. m. braun , a. lenz , m. wittmann , phys .
d * 73 * , 094079 ( 2006 ) .
i. v. anikin , v. m. braun , anf n. offen , phys .
d * 88 * , 114021 ( 2013 ) .
* fig . ( 1 ) * the dependence of the helicity amplitude @xmath90 on @xmath4 at @xmath93 , and at several fixed values of the borel mass parameter @xmath55 .
+ + * fig .
( 2 ) * the same as in fig . ( 1 ) , but for the helicity amplitude @xmath91 .
+ + * fig .
( 3 ) * the same as in fig . ( 1 ) , but for the helicity amplitude @xmath92 . | the @xmath0 reaction is studied in framework of the light cone qcd sum rules method .
the sum rules for the multipole form factors , namely , for the magnetic dipole @xmath1 , electric quadrupole @xmath2 , and coulomb quadrupole @xmath3 , form factors responsible for for the @xmath0 transition are derived . using these sum rules ,
the helicity amplitudes are presented as the combinations of these multipole form factors , and their @xmath4 dependences are studied .
we compare our results on the helicity amplitudes with those of the spectator model predictions , and with the existing experimental data . # 1#2#3 @xmath5 #
1#2#3 @xmath5 0^*0 5_5 o _ ^0 _ pacs numbers : 11.55.hx , 13.30.ce , 13.40.gp , 14.20.jn |
little is currently known about the detailed geometry and structure of the broad - line region ( blr ) in quasi - stellar objects ( qsos ) due to the difficulty in spatially resolving the central regions even in the most nearby active galactic nuclei ( agn ) , the limited sensitivity of photo - ionization processes to the detailed blr geometry , and to the complex nature of reverberation transfer functions ( peterson 1997 ; netzer & peterson 1997 ) .
for example , it is unclear whether the blr gas has a spherical or disk - like distribution , and whether the blr consists of discrete clouds or is part of a continuous gas distribution , such as an accretion disk wind [ boroson & green 1992 ( bg92 ) ; murray & chiang 1997 ] . an improved knowledge of the geometry and structure of the centrally emitting and absorbing gas regions and their relation to one another and to the continuum source(s ) may help us to better understand the observed continuum - line and line - line correlations ( bg92 ; brotherton 1996 ; wilkes 1999 ; wills 1999 ) , and furthermore shed light on the interrelation of individual agn subclasses . the best constraints on the geometry and structure have so far been obtained from reverberation mapping ( see the excellent reviews by peterson 1993 ; netzer & peterson 1997 ) of low - luminosity agn showing ( 1 ) the blr is rather compact ( the outer radius , r@xmath3 is smaller than 0.3 pc @xmath4arcsec ) , but is geometrically thick with r@xmath5 10 r@xmath6 , ( 2 ) a certain degree of ionization stratification is present in the blr gas , and ( 3 ) the blr size and the ( bolometric ) luminosity of the active nucleus are correlated ( kaspi 2000 ) .
some clues to the relative position of the continuum source , and the scattering , emitting , and absorbing components , respectively , can be obtained from polarization studies of broad - absorption line qsos ( ogle 1997 ) which argue that they are highly inclined `` radio - quiet '' qsos .
additional and independent clues to the blr gas kinematics may be obtained by studying the widths of the broad emission lines themselves .
the gravitational potential of the central engine is thought to dominate the kinematics of the blr gas and the line width ( fwhm ) to estimate the gravitational potential at the location of the gas , ( peterson & wandel 1999 ) .
if the motion of the blr gas has a preferred direction , as opposed to disorganized virial motions around the central black hole , we would expect the projection of this velocity field on our sight line to vary as the source axis inclination varies in the agn population . for a sample of broad - line radio sources wills & browne ( 1986 ) found fwhm(h@xmath7 ) to relate to the radio core - fraction , @xmath8 , an estimator of the radio axis inclination to our line of sight , such that face - on sources lack very broad lines .
they conclude that h@xmath7 is emitted by a disk configuration perpendicular to the radio axis , which offers support to the two - component model for the high and low ionization lines ( collin - souffrin 1988 ) .
the low - ionization , very high - density environment of the accretion disk fits the requirements of the physical conditions in the balmer line emitting gas .
the low level of long - term variability in the broad h@xmath7 wings ( maoz 1994 ; peterson 2000 ) seems to support the disk interpretation [ corbin & smith 2000 ; but see brotherton ( 1996 ) for counter arguments ] . also , zheng ( 1992 ) and corbin ( 1997 ) model the high - velocity blr emission finding consistency with keplerian motion in low-@xmath9 sources .
collin - souffrin suggested a polar outflow origin for the high - ionization lines ( hils ) , thereby explaining their observed peak blueshift relative to the low - ionization lines ( lils ; gaskell 1982 ; wilkes 1984 ; tytler & fan 1992 ) .
however , in the absence of a definitive model a spherical distribution of dense , discrete hil - emitting clouds is often adopted in blr studies to explain trends such as the inverse correlation between [ oiii]@xmath05007 and the feii emission ( bg92 ) . in order to explore these possible models we studied the inclination dependence of the widths of civ@xmath01549 , representative of the hils , relative to those reported for h@xmath7 in a sample of radio - powerful qsos ( rlqs ) .
we find interesting constraints applicable to the hil geometry , which we present below .
this project is part of an extended study of the ultraviolet ( uv ) broad emission lines in moderate-@xmath9 qsos with a range of radio power .
the details of the sample and the optical data of the extended study are described by vestergaard ( in preparation , hereafter paper 1 ) .
we summarize the main points below .
the radio data are described by barthel ( 1988 ) , lonsdale , barthel & miley ( 1993 ) and barthel , vestergaard & lonsdale ( 2000 ) .
the current sample consists of 25 lobe - dominated ( ldqs ; r@xmath10 0.5 ; r@xmath11 is defined below ) and 11 core - dominated ( cdqs ; r@xmath12 0.5 ) rlqs , chosen from the larger sample of @xmath1 2 rlqs , originally selected from barthel , tytler & thomson ( 1990 ) , and the 3c and 4c catalogues . to ensure that we study the inclination dependence in the line widths , and not , say , a dependence on intrinsic source brightness
, the cdqs and ldqs were selected such that their distributions of extended 5ghz radio luminosity , l@xmath13 , are similar .
a kolmogorov - smirnov test confirms this to a confidence level of 97.7% .
the l@xmath13-ranges and the mean and median values also match to within the standard error , while the m@xmath14-ranges match to within the expected uncertainty of @xmath150.5@xmath16 ( table 1 ) .
three of the ldqs are compact ( @xmath1715kpc ) but their @xmath18-values could be measured from our high - resolution radio images .
highly variable rlqs are excluded . for the rlqs not presented by barthel ( 1990 ) we obtained mmt blue channel spectra during 1994 october
1996 november .
the rlqs were observed at airmasses below 1.4 and with a 1 slit placed at the parallactic angle .
the barthel data were not well flux - calibrated and so were recalibrated using short mmt exposures .
spurious noise spikes and narrow absorption lines superposed on the emission lines were corrected by generating smooth fits to the profiles .
the multiple gaussian fitting ( laor 1994 ; paper 1 ) that was applied to the civ
heii line complex also allows a deblending of the line emission contributions , especially in the red wing of civ , which in a few objects shows signs of weak feii emission .
this is not expected to affect the current results adversely .
the limitations imposed by the feii emission will be discussed in paper 1 .
the profile parameters were measured on the smooth civ fits to ensure that they are measured consistently , objectively , and with minimum uncertainties .
the full line widths at fractional profile heights , y (= 10% , 20% , ... , 90% ) were determined ( fwym ) , along with the interpercentile velocity widths , ipv(y% ) ( whittle 1985 ) .
reliable error estimates of each line parameter , based on continuum and flux uncertainties , were determined in a similar fashion as outlined by whittle ( 1985 ; see paper 1 for details ) .
the radio core - fraction is defined here as @xmath18 = s@xmath19/s@xmath20 , where s@xmath21@xmath15@xmath22 are the observed fluxes at 5ghz measured from our vla images ; the spectral index , @xmath23 , is defined as @xmath24 ( or @xmath25 if the former is unavailable ) . a k - correction
= 50 km s@xmath26 mpc@xmath26 , q@xmath27 = 0 is used throughout ] was applied to @xmath18 to permit a comparison in the rest - frame : @xmath28 $ ] , where @xmath290 and @xmath300.8 ( kellerman & pauliny - toth 1981 ) .
our definition of @xmath18 is less sensitive to variations in the density of the environment of the radio source than the @xmath2 defined as @xmath31 , especially for face - on objects , where @xmath32 tends to be relatively weak .
we also consider the inclination measure , @xmath33 , defined as the ratio of the 5ghz core flux density to the visual flux density with applied k - correction ( wills & brotherton 1995 ) .
it is also less sensitive to the density of the radio source environment .
fwhm(civ ) shows no relation with @xmath34 ( the confidence level p , where p is the probability of no correlation ] [ herafter cl ] of the spearman rank correlation test is @xmath1579% ) .
however , fw20m(civ ) , representative of the width of the broad civ base , shows a strong inverse correlation with @xmath34 at the 99.7% cl ( fig .
[ fig1.fig]a ) .
support for the reality of this effect comes from the similar relations for fw10 m , fw30 m , and fw40 m ( cls : 97.5% , 99.6% , and 98.6% ) . as an `` area parameter '' , ipv(y )
is much less sensitive to the varying line - core strength which causes scatter in the fwym measures ( paper 1 ) .
the fact that ipv(4070% ) show an inverse correlation with @xmath34 at the 9798% cl , also supports the reality of the fw20 m trend .
the ipv(y@xmath3540% ) is more sensitive to variations in the extreme wings than fwym , evident from the larger scatter and errors .
this does not jeopardize the above results for ipv(4070% ) [ paper 1 ] .
the @xmath33 measure shows similar relations as @xmath34 though with relatively lower significance .
this is most likely due to the fact that a _ zone of avoidance _ is more evident in these relations ( figure 1b ) though it is less apparent than that presented for h@xmath7 by wills & browne ( 1986 ) .
brotherton ( 1996 ) , however , finds clear anti - correlations for fw25 m and fwhm of h@xmath7 ( but none for fw75 m ) .
the civ fw20 - 30 m relations with core - dominance must primarily be driven by inclination and not some other property because : * ( 1 ) * the l@xmath13 distributions are similar for ldqs and cdqs .
this is further confirmed by the at most weak fw20 - 30 m
l@xmath13 relations ( cl @xmath3695% ) , * ( 2 ) * the lack of ew correlations with core - dominance ( cls @xmath3592% ) indicates that the ( luminosity normalized ) line flux does not change with inclination , * ( 3 ) * the lack of fw70 - 90 m relations with core - dominance ( cls @xmath3537% ) shows there is no weakening of the narrow line core in ldqs which could yield an apparent but unreal fw20 - 30 m inclination dependence , and * ( 4 ) * no correlation with inclination is found for m@xmath14 , l@xmath13 , and redshift .
the line width relations with inclination directly show that the high - velocity field projected along the line of sight increases with source inclination .
the velocity field of blr gas ( in discrete clouds or a continuous distribution ) in organized orbits around the source axis is expected to behave this way .
the similar h@xmath7 and civ relations imply that , to first order , the kinematical structures of the two high - velocity line - emitting regions , and by inference those of hils and lils , are similar . item ( 2 ) above also implies that the line and continuum fluxes are anisotropically emitted to about the same degree .
highly beamed sources are excluded and the uv continuum is also not expected to be significantly beamed ( wills & brotherton 1995 ) . item ( 2 ) confirms this .
though these results do not provide independent evidence for a disk - like geometry , they are consistent with the high - velocity hil gas being physically close to the accretion disk and with its predominant motion being in a plane parallel to the disk
. however , the civ emission can not originate in the low - ionization accretion disk - material itself , while the h@xmath7 emission may ( collin - souffrin 1988 ) .
we therefore propose that the high - velocity hil gas is planar and lies in close proximity to the disk , and with similar overall kinematics . in our ongoing work
we will re - address this fw20 - 30 m inclination dependence issue with larger and optimally constructed l@xmath13-samples of ldqs and cdqs .
our results argue against two possible blr geometries : ( i ) the velocity field of a spherical distribution of randomly orbiting clouds is not expected to vary with inclination , and ( ii ) a polar outflow of the hil gas would show the opposite inclination dependence to that seen here for civ .
however , a planar hil velocity - field and geometry is consistent with a disk - wind , where the broad lines are emitted at the base of dense , optically thick winds evaporating off the accretion disk ( murray & chiang 1997 , 1998 ) .
this model can successfully explain a number of observational details .
bottorff ( 1997 ) and knigl & kartje ( 1994 ) present alternative disk - like blr models . item ( 3 ) above implies that the lowest - velocity blr gas has a random velocity field relative to inclination .
it may not be physically associated with the high - velocity gas located close to the accretion disk .
low - density hil gas of lower optical depth at higher disk - altitudes , similar to the proposed intermediate line region ( ilr ; brotherton 1994 ) , fits this requirement ; the altitude is low enough to maintain the high ionization level . in the murray & chiang ( 1997 ) disk - wind model
the narrow line - core is emitted by the wind at much larger disk - radii than the broad line - base .
this predicts similar inclination dependences for the high and low velocity fields in contrast to our results and so requires a modification of the model , possibly along the lines suggested above .
brotherton ( 1996 ) finds very similar results for the inclination dependence of the high and low velocity h@xmath7 emission in a sample of rlqs .
the vblr / ilr picture , which he advocates , is not geometry specific , but is fully consistent with the results and interpretations presented here .
a disk - like geometry of the hil gas predicts : * ( 1 ) * other hils ( ly@xmath23 , nv , siiv+oiv ] , and heii ) should show similar dependences on source inclination ; such a study is underway , * ( 2 ) * the low - velocity dispersion ( @xmath36 2000 ^-1 kms@xmath26 ) of the narrowest line core of the hils should be independent of inclination , assuming this gas is located at relatively high disk - altitudes , and * ( 3 ) * the civ and h@xmath7 profile parameters should behave similarly with varying source inclination , excluding the civ narrow line core and the contribution to h@xmath7 from the narrow line region .
no detailed study exists of high - redshift qsos ; the fwhms of civ and h@xmath7 correlate somewhat at low redshift ( wilkes 1999 ) .
we are currently studying this issue for low - redshift qsos , based in part on _ hubble space telescope _ archival data .
despite years of study little is known for certain about the geometry of the blr of agn beyond its compact size ( @xmath36light - weeks ) .
the present results indicate that the high - velocity hil gas is located in a disk - like configuration close to the accretion disk , long thought to emit the lils , and with similar kinematics .
our results argue against a spherical cloud distribution with random cloud velocities and a high - ionization polar outflow .
if confirmed , the implications are intriguing ; they may help to constrain interpretations of blr data and to suggest new studies to shed light on the detailed structure and kinematics of the agn blr .
-0.2 cm we thank drs .
smita mathur and brad peterson for comments on this paper , dr . bev wills for useful discussions of the @xmath33 measure , and dr .
norm murray of the disk - wind model .
mv is very pleased to thank the smithsonian astrophysical observatory for their hospitality and gratefully acknowledges financial support from the columbus fellowship at the ohio state university , the danish natural sciences research council ( snf-9300575 ) , the danish research academy ( dfa - s930201 ) and a research assistantship at smithsonian astrophysical observatory through nasa grants nagw-4266 , nagw-3134 , and nag5 - 4089 to bjw .
bjw gratefully acknowledges financial support from nasa contract nas 8 - 39073 ( _ chandra _ x - ray center ) .
pdb acknowledges contributions by drs .
david tytler and george miley in the early stages of this project , and travel support by the leids kerkhoven - bosscha fonds .
barthel , p. d. , miley , g. k. , schilizzi , r. t. , & lonsdale , c. j. 1988 , , 73 , 515 barthel , p. d. , tytler , d. r. , & thomson , b. 1990 , , 82 , 339 barthel , p. d. , vestergaard , m. , & lonsdale , c. j. 2000 , , 354 , 7 boroson , t. a. , & green , r. f. 1992 , , 80 , 109 ( bg92 ) bottorff , m , korista , k. t. , shlosman , i. , blandford , r. d. 1997 , , 479 , 200 brotherton , m. s. , wills , b. j. , francis , p. j. , steidel , c. c. 1994 , , 430 , 495 brotherton , m. s. 1996 , , 102 , 1 collin - souffrin , s. , dyson , j. e. , mcdowell , j. c. , & perry , j. j. 1988 , , 232 , 539 corbin , m. r. 1997 , , 485 , 517 corbin , m. r. & smith , p. s. 2000 , , 532 , 136 gaskell , c. m. 1982 , , 263 , 79 kaspi , s. , smith , p. s. , netzer , h. , maoz , d. , jannuzi , b. t. , & giveon , u. 2000 , , 533 , 631 kellermann , k.i . , pauliny - toth , i. i. k. 1981 , , 19 , 373 knigl , a. , & kartje , j. f. 1994 , , 434 , 446 laor , a. , bahcall , j. n. , jannuzi , b. t. , schneider , d. p. , green , r. f. , & hartig , g. f. 1994 , , 420 , 110 lonsdale , c. j. , barthel , p. d. , & miley , g. k. 1993 , , 87 , 63 maoz , d. , smith , p. s. , jannuzi , b. t. , kaspi , s. , & netzer , h. 1994 , , 421 , 34 murray , n. , & chiang , j. 1997 , , 474 , 91 murray , n. , & chiang , j. 1998 , , 494 , 125 netzer , h. , & peterson , b. m. 1997 , in astronomical time series , ed .
d. maoz , a. sternberg , & e. m. leibowitz ( dordrecht : kluwer ) , 85 ogle , p. m. 1997 , in asp conf .
128 , mass ejection from active galactic nuclei , ed .
n. arav , i. shlosman , & r. j. weymann ( san francisco : asp ) , 78 peterson , b.m .
1993 , , 105 , 247 peterson , b.m .
1997 , an introduction to active galactic nuclei ( cambridge : cup ) peterson , b. m. , & wandel , a. 1999 , , 521 , l95 peterson , b. m. et al .
2000 , , in press , astro - ph/0005433 tytler , d. , & fan , x .-
1992 , , 79,1 whittle , m. 1985 , , 213 , 1 wilkes , b. j. 1984 , , 207 , 73 wilkes , b. j. , kuraszkiewicz , j. , green , p. j. , mathur , s. , mcdowell , j. c. 1999 , , 513 , 76 wills , b. j. , laor , a. , brotherton , m. s. , wills , d. , wilkes , b. j. , ferland , g. j. , shang , z. 1999 , , 515 , l53 wills , b. j. , & brotherton , m. s. 1995 , , 448 , l81 wills , b. j. , & browne , i. w. a. 1986 , , 302 , 56 zheng , w. 1992 , , 385 , 127 | we present evidence that the high - velocity civ@xmath01549 emission line gas of radio - loud quasars may originate in a disk - like configuration , in close proximity to the accretion disk often assumed to emit the low - ionization lines . for a sample of 36 radio - loud @xmath12 quasars
we find the 2030% peak width to show significant inverse correlations with the fractional radio core - flux density , @xmath2 , the radio axis inclination indicator .
highly inclined systems have broader line wings , consistent with a high - velocity field perpendicular to the radio axis .
by contrast , the narrow line - core shows no such relation with @xmath2 , so the lowest velocity civ - emitting gas has an inclination independent velocity field . we propose that this low - velocity gas is located at higher disk - altitudes than the high - velocity gas .
a planar origin of the high - velocity civ - emission is consistent with the current results and with an accretion disk - wind emitting the broad lines .
a spherical distribution of randomly orbiting broad - line clouds and a polar high - ionization outflow are ruled out . |
the two nearby galaxies m31 and m33 , are stepping stones to most of our current effort to understand the evolving universe at large scales . first , they are essential to the calibration of the extragalactic distance scale ( jacoby et al . 1992 ; tonry et al . 1997 ) .
second , they constrain population synthesis models for early galaxy formation and evolution , and provide the stellar luminosity calibration .
there is one simple requirement for all this accurate distances .
detached eclipsing binaries ( debs ) have the potential to establish distances to m31 and m33 with an unprecedented accuracy of better than 5% and possibly to better than 1% .
these distances are now known to no better than 10 - 15% , as there are discrepancies of @xmath1 between rr lyrae and cepheids distance indicators ( e.g. huterer , sasselov & schechter 1995 ) .
detached eclipsing binaries ( for reviews see andersen 1991 , paczyski 1997 ) offer a single step distance determination to nearby galaxies and may therefore provide an accurate zero point calibration a major step towards very accurate determination of the hubble constant , presently an important but daunting problem for astrophysicists ( see the papers from the recent `` debate on the scale of the universe '' : tammann 1996 , van den bergh 1996 ) .
the detached eclipsing binaries have yet to be used ( huterer et al .
1995 ; hilditch 1996 ) as distance indicators to m31 and m33 .
according to hilditch ( 1996 ) , there are about 60 eclipsing binaries of all kinds known in m31 ( gaposchkin 1962 ; baade & swope 1963 ; baade & swope 1965 ) and only _ one _ in m33 ( hubble 1929 ) !
only now does the availability of large format ccd detectors and inexpensive cpus make it possible to organize a massive search for periodic variables , which will produce a handful of good deb candidates .
these can then be spectroscopically followed - up with the powerful new 6.5 - 10 meter telescopes .
the study of cepheids in m31 and m33 has a venerable history ( hubble 1926 , 1929 ; gaposchkin 1962 ; baade & swope 1963 ; baade & swope 1965 ) . in the 80 s freedman & madore ( 1990 ) and freedman ,
wilson & madore ( 1991 ) studied small samples of the earlier discovered cepheids , to build pl relations in m31 and m33 , respectively .
however , both the sparse photometry and the small samples do not provide a good basis for obtaining direct baade - wesselink distances ( e.g. krockenberger , sasselov & noyes 1997 ) to cepheids the need for new digital photometry has been long overdue .
recently , magnier et al .
( 1997 ) surveyed large portions of m31 , which have previously been ignored , and found some 130 new cepheid variable candidates .
their light curves are however rather sparsely sampled and in @xmath2 band only . in kaluzny
et al . ( 1998 ) ( hereafter : paper i ) , the first paper of the series , we presented the catalog of variable stars found in one of the fields in m31 , called m31b . here
we present the catalog of variables from the neighboring field m31a . in sec.2
we discuss the selection of the fields in m31 and the observations . in sec.3
we describe the data reduction and calibration . in sec.4
we discuss briefly the automatic selection we used for finding the variable stars . in sec.5
we discuss the classification of the variables . in sec.6
we present the catalog of variable stars .
m31 was primarily observed with the mcgraw - hill 1.3-meter telescope at the mdm observatory .
we used the front - illuminated , loral @xmath3 ccd wilbur ( metzger , tonry & luppino 1993 ) , which at the @xmath4 station of the 1.3-meter has a pixel scale of @xmath5 and field of view of roughly @xmath6 .
we used kitt peak johnson - cousins @xmath7 filters .
some data for m31 were also obtained with the 1.2-meter telescope at the flwo , where we used `` andycam '' with thinned , back - side illuminated , ar coated loral @xmath3 ccd . the pixel scale happens to be essentially the same as at the mdm 1.3-meter telescope .
we used standard johnson - cousins @xmath7 filters .
fields in m31 were selected using the mit photometric survey of m31 by magnier et al .
( 1992 ) and haiman et al .
( 1994 ) ( see paper i , fig.1 ) .
we selected six @xmath8 fields , m31a f , four of them ( a d ) concentrated on the rich spiral arm in the north - eastern part of m31 , one ( e ) coinciding with the region of m31 searched for microlensing by crotts & tomaney ( 1996 ) , and one ( f ) containing the giant star formation region known as ngc206 ( observed by baade & swope 1963 ) . fields a c were observed during september and october 1996 58 times per night in the @xmath2 band , resulting in total of 110 - 160 @xmath2 exposures per field .
fields d f were observed once a night in the @xmath2-band . some exposures in @xmath9 and @xmath10 bands
were also taken .
m31 was also occasionally observed at the flwo 1.2-meter telescope , whose main target was m33 . in this paper
we present the results for the m31a field .
we obtained for this field useful data during 29 nights at the mdm , collecting a total of 109 exposures of @xmath11 in @xmath2 , 27 exposures of @xmath12 in @xmath10 and 2 exposures of @xmath13 in @xmath9 .
we also obtained for this field useful data during 15 nights at the flwo , collecting a total of 8 exposures of @xmath11 in @xmath2 and 18 exposures of @xmath12 in @xmath10 .
the complete list of exposures for this field and related data files are available through anonymous ftp on cfa-ftp.harvard.edu , in pub / kstanek / direct directory .
please retrieve the readme file for instructions .
additional information on the direct project is available through the www at http://cfa - www.harvard.edu/~kstanek / direct/.
the details of the reduction procedure are given in paper i. preliminary processing of the ccd frames was done with the standard routines in the iraf - ccdproc package .
stellar profile photometry was extracted using the _ daophot / allstar _ package ( stetson 1987 , 1991 ) .
we selected a `` template '' frame for each filter using a single frame of particularly good quality .
these template images were reduced in a standard way ( paper i ) .
other images were reduced using _
allstar _ in the fixed - position - mode using as an input the transformed object list from the template frames . for each frame
the list of instrumental photometry derived for a given frame was transformed to the common instrumental system of the appropriate `` template '' image .
photometry obtained for the @xmath2 & @xmath10 filters was combined into separate data bases . unlike for the m31b field , m31a images obtained at the flwo were reduced using mdm `` templates '' . to obtain the photometric calibration we observed 4 landolt ( 1992 ) fields containing a total of 18 standards stars .
these fields were observed through @xmath7 filters at air - masses ranging from 1.2 to 1.70 .
the transformation from the instrumental to the standard system was derived and described in paper i. the derived transformation satisfactorily reproduces the @xmath2 magnitudes and @xmath14 colors .
the @xmath15 transformation reproduces the standard system poorly , and we decided to drop the @xmath9 data from our analysis , especially since we took only 2 @xmath9 frames . to check the internal consistency of our photometry we compared the photometry for 20 @xmath16 and 47 @xmath17 common stars in the overlap region between the fields m31a and m31b ( paper i , fig.1 )
there was an offset of @xmath18 in @xmath2 and @xmath19 in @xmath10 , i.e. well within our estimate of the @xmath20 systematic error discussed in paper i. we also derived equatorial coordinates for all objects included in the data bases for the @xmath2 filter .
the transformation from rectangular coordinates to equatorial coordinates was derived using @xmath21 stars identified in the list published by magnier et al .
the procedure for selecting the variables was described in detail in paper i , so here we only give a short description , noting changes when necessary .
the reduction procedure described in previous section produces databases of calibrated @xmath2 and @xmath10 magnitudes and their standard errors .
the @xmath2 database for m31a field contains 10084 stars with up to 117 measurements , and the @xmath10 database contains 21341 stars with up to 45 measurements .
fig.[fig : dist ] shows the distributions of stars as a function of mean @xmath22 or @xmath23 magnitude .
as can be seen from the shape of the histograms , our completeness starts to drop rapidly at about @xmath24 and @xmath25 .
the primary reason for this difference in the depth of the photometry between @xmath2 and @xmath10 is the level of the combined sky and background light , which is about three times higher in the @xmath10 filter than in the @xmath2 filter .
the measurements flagged as `` bad '' and measurements with errors exceeding the average error by more than @xmath26 are removed ( paper i ) .
usually 010 points are removed , leaving the majority of stars with roughly @xmath27 @xmath2 measurements . for further analysis
we use only those stars which have at least @xmath28 measurements .
there are 8521 such stars in the @xmath2 database of the m31a field .
our next goal is to select objectively a sample of variable stars from the total sample defined above .
there are many ways to proceed , and we largely follow the approach of stetson ( 1996 ) .
the procedure is described in more detail in paper i. in short , for each star we compute the stetson s variability index @xmath29 ( paper i , eq.7 ) , and stars with values exceeding some minimum value @xmath30 are considered candidate variables .
the definition of stetson s variability index includes the standard errors of individual observations .
if , for some reason , these errors were over- or underestimated , we would either miss real variables , or select spurious variables as real ones .
using the procedure described in paper i , we scale the _ daophot _ errors to better represent the `` true '' photometric errors .
we then select the candidate variable stars by computing the value of @xmath29 for the stars in our @xmath2 database .
we used a cutoff of @xmath31 and additional cuts described in paper i to select 183 candidate variable stars ( about 2% of the total number of 8521 ) . in fig.[fig : stetj ] we plot the variability index @xmath29 vs. apparent visual magnitude @xmath22 for 8521 stars with @xmath32 .
we based our candidate variables selection on the @xmath2 band data collected at the mdm and the flwo telescopes .
we also have the @xmath10 band data for the field , up to 27 mdm epochs and up to 18 flwo epochs . as mentioned above
, the @xmath10 photometry is not as deep as the @xmath2 photometry , so some of the candidate variable stars do not have an @xmath10 counterpart .
we will therefore not use the @xmath10 data for the period determination and broad classification of the variables .
we will however use the @xmath10 data for the `` final '' classification of some variables .
next we searched for the periodicities for all 183 candidate variables , using a variant of the lafler - kinman ( 1965 ) technique proposed by stetson ( 1996 ) .
starting with the minimum period of @xmath33 , successive trial periods are chosen so @xmath34 where @xmath35 is the time span of the series .
the maximum period considered is @xmath36 . for each candidate
variable 10 best trial periods are selected ( paper i ) and then used in our classification scheme .
the variables we are most interested in are cepheids and eclipsing binaries ( ebs ) .
we therefore searched our sample of variable stars for these two classes of variables .
as mentioned before , for the broad classification of variables we restricted ourselves to the @xmath2 band data .
we will , however , present and use the @xmath10 band data , when available , when discussing some of the individual variable stars . in the search for cepheids we followed the approach by stetson ( 1996 ) of fitting template light curves to the data .
we used the parameterization of cepheid light curves in the @xmath2 band as given by stetson ( 1996 ) . unlike for the m31b field , we classified the star as a cepheid if the reduced @xmath37 of the fit was factor of 3 smaller than the reduced @xmath37 of a straight line fit , not a factor of 2 smaller .
we also allowed for periods longer than @xmath38 , not longer than @xmath39 as in paper i , which resulted in finding two possible cepheids with periods @xmath40 .
there was a total of 51 variables passing all of the criteria .
their parameters and light curves are presented in the sections 6.2 , 6.3 .
for eclipsing binaries ( ebs ) we used similar search strategy to that described in detail in paper i , but we simplified the condition for the reduced @xmath37 of the fit to be at least 1.75 smaller than the reduced @xmath37 of a straight line fit . within our assumption the light curve of an eb is determined by nine parameters : the period , the zero point of the phase , the eccentricity , the longitude of periastron , the radii of the two stars relative to the binary separation , the inclination angle , the fraction of light coming from the bigger star and the uneclipsed magnitude .
a total of 15 variables passing all of the criteria and their parameters and light curves are presented in the section 6.1 .
after we selected 15 eclipsing binaries and 51 possible cepheids , we were left with 118 `` other '' variable stars . after raising the threshold of the variability index to @xmath41 ( paper i )
we are left with 29 variables , which we preliminary classify as `` miscellaneous '' .
we then go back to the ccd frames and try to see by eye if the inferred variability is indeed there , especially in cases when the light curve is very noisy / chaotic .
we decided to remove 2 dubious cepheids and 18 dubious miscellaneous variables from the sample , which leaves 11 variables which we classify as miscellaneous .
their parameters and light curves are presented in the section 6.4 .
in this section we present light curves and some discussion of the 75 variable stars discovered in our survey .
complete @xmath2 and ( when available ) @xmath10 photometry and @xmath42 ( @xmath43 ) @xmath2 finding charts for all variables are available through the anonymous ftp on cfa-ftp.harvard.edu , in pub / kstanek / direct directory .
please retrieve the readme file for the instructions and the list of files .
these data can also be accessed through the www at the http://cfa - www.harvard.edu/~kstanek / direct/. the ccd frames for this field , and also for the m31b field , are available on request from k. z. stanek ( [email protected] ) .
the variable stars are named according to the following convention : letter v for `` variable '' , the number of the star in the @xmath2 database , then the letter `` d '' for our project , direct , followed by the name of the field , in this case ( m)31a , e.g. v3407 d31a .
tables [ table : ecl ] , [ table : ceph ] , [ table : per ] and [ table : misc ] list the variable stars sorted broadly by four categories : eclipsing binaries , cepheids , other periodic variables and `` miscellaneous '' variables , in our case meaning `` variables with no clear periodicity '' .
some of the variables which were found independently by survey of magnier et al .
( 1997 ) are denoted in the `` comments '' by `` ma97 i d '' , where the `` i d '' is the identification number assigned by magnier at al .
we also identify several variables found by us in paper i. note that this is a first step in a long - term project and we are planning to collect additional data and information of various kind for this and other fields . as a result , the current catalog might undergo changes , due to addition , deletion or re - classification of some variables . in table [ table : ecl ] we present the parameters of the 15 eclipsing binaries in the m31a field .
the lightcurves of these variables are shown in figs.[fig : ecl1][fig : ecl3 ] , along with the simple eclipsing binary models discussed in the paper i. the variables are sorted in the table [ table : ecl ] by the increasing value of the period @xmath44 . for each eclipsing binary
we present its name , 2000.0 coordinates ( in degrees ) , value of the variability index @xmath29 , period @xmath44 , magnitudes @xmath45 and @xmath46 of the system outside of the eclipse , and the radii of the binary components @xmath47 in the units of the orbital separation .
we also give the inclination angle of the binary orbit to the line of sight @xmath48 and the excentricity of the orbit @xmath49 .
the reader should bear in mind that the values of @xmath50 and @xmath49 are derived with a straightforward model of the eclipsing system , so they should be treated only as reasonable estimates of the `` true '' value .
one of the eclipsing binaries found , v4636 d31a , is a good deb candidate , with reasonably deep eclipses and the ellipticity indicating that the system is young and unevolved .
however , much better light curve is necessary to accurately establish the properties of the system . in table [ table : ceph ] we present the parameters of 43 cepheids in the m31a field , sorted by the period @xmath44 . for each cepheid
we present its name , 2000.0 coordinates , value of the variability index @xmath29 , period @xmath44 , flux - weighted mean magnitudes @xmath51 and ( when available ) @xmath52 , and the @xmath2-band amplitude of the variation @xmath53 . in figs.[fig : ceph1][fig : ceph8 ] we show the phased @xmath54 lightcurves of our cepheids . also shown is the best fit template lightcurve ( stetson 1996 ) , which was fitted to the @xmath2 data and then for the @xmath10 data only the zero - point offset was allowed . for some of the variables preliminary classified as cepheids ( section 5.3.1 ) , we decided upon closer examination to classify them as `` other periodic variables '' . in table
[ table : per ] we present the parameters of 6 possible periodic variables other than cepheids and eclipsing binaries in the m31a field , sorted by the increasing period @xmath44 . for each variable
we present its name , 2000.0 coordinates , value of the variability index @xmath29 , period @xmath44 , error - weighted mean magnitudes @xmath22 and ( when available ) @xmath23 . to quantify the amplitude of the variability
, we also give the standard deviations of the measurements in the @xmath2 and @xmath10 bands , @xmath55 and @xmath56 .
lrrrrrrrrcrc v1555 & 11.2717 & 41.6462 & 1.07 & 0.917 & 20.64 & & 0.59 & 0.41 & 76 & 0.00 & v6913 d31b v9936 & 11.4523 & 41.7306 & 0.77 & 0.930 & 21.33 & & 0.65 & 0.34 & 61 & 0.06 & v4741 & 11.3337 & 41.7511 & 1.43 & 1.604 & 20.35 & 20.16 & 0.51 & 0.40 & 73 & 0.03 & v8420 & 11.4072 & 41.7188 & 1.21 & 1.931 & 20.71 & 20.58 & 0.42 & 0.37 & 74 & 0.04 & v6024 & 11.3611 & 41.6841 & 0.83 & 2.083 & 20.93 & & 0.38 & 0.37 & 69 & 0.00 & v7393 & 11.3851 & 41.7951 & 0.85 & 2.786 & 20.14 & & 0.35 & 0.31 & 80 & 0.03 & v6450 & 11.3681 & 41.7400 & 1.39 & 3.076 & 20.60 & 20.63 & 0.56 & 0.44 & 67 & 0.04 & v6527 & 11.3671 & 41.8256 & 0.84 & 4.180 & 20.77 & & 0.31 & 0.31 & 76 & 0.01 & v5912 & 11.3563 & 41.7510 & 0.88 & 5.006 & 19.92 & & 0.67 & 0.33 & 61 & 0.01 & v9840 & 11.4495 & 41.7034 & 0.78 & 5.215 & 21.51 & & 0.63 & 0.36 & 75 & 0.03 & v8153 & 11.4026 & 41.6701 & 0.87 & 5.794 & 21.49 & 20.36 & 0.56 & 0.44 & 83 & 0.14 & v5407 & 11.3464 & 41.7504 & 1.02 & 7.061 & 20.33 & 19.51 & 0.51 & 0.44 & 57 & 0.02 & v538 & 11.2420 & 41.6695 & 0.85 & 7.171 & 21.07 & 20.44 & 0.59 & 0.40 & 66 & 0.00 & v4636 & 11.3321 & 41.7514 & 0.80 & 8.181 & 19.44 & 19.43 & 0.14 & 0.14 & 83 & 0.18 & deb v6423 & 11.3671 & 41.7533 & 0.85 & 11.782 & 20.80 & 19.89 & 0.54 & 0.46 & 57 & 0.15 & [ table : ecl ] lrrrrrrrc v3885 & 11.3176 & 41.7868 & 1.13 & 3.801 & 21.87 & 21.63 & 0.34 & v4585 & 11.3337 & 41.6685 & 1.18 & 3.953 & 22.06 & 21.71 & 0.44 & v8435 & 11.4083 & 41.6933 & 1.05 & 4.097 & 22.05 & 21.37 & 0.43 & v3071 & 11.3036 & 41.6866 & 1.05 & 4.557 & 21.49 & & 0.23 & v8798 & 11.4135 & 41.8081 & 1.36 & 4.861 & 22.09 & & 0.38 & v9833 & 11.4461 & 41.7914 & 1.60 & 4.867 & 21.22 & 20.20 & 0.29 & v9531 & 11.4361 & 41.7111 & 1.69 & 4.869 & 21.55 & 20.55 & 0.36 & v7466 & 11.3867 & 41.7830 & 1.15 & 5.007 & 21.42 & 20.71 & 0.29 & v3584 & 11.3125 & 41.7674 & 0.80 & 5.020 & 21.09 & 20.12 & 0.14 & v6800 & 11.3780 & 41.6476 & 0.91 & 5.077 & 21.82 & & 0.35 & v5348 & 11.3479 & 41.6677 & 1.65 & 5.346 & 21.61 & 20.75 & 0.41 & v8573 & 11.4116 & 41.7128 & 1.35 & 5.476 & 21.39 & 20.50 & 0.32 & ma97 125 v8232 & 11.4011 & 41.7727 & 1.37 & 5.681 & 21.17 & 19.88 & 0.27 & ma97 124 v589 & 11.2430 & 41.6840 & 2.42 & 6.192 & 21.25 & 20.25 & 0.40 & v3142 & 11.3047 & 41.7006 & 1.83 & 6.244 & 21.52 & 21.09 & 0.38 & v2770 & 11.2952 & 41.7592 & 1.62 & 6.413 & 21.25 & 20.46 & 0.29 & v6842 & 11.3770 & 41.7058 & 1.20 & 6.482 & 21.65 & 20.79 & 0.30 & v9473 & 11.4337 & 41.7366 & 1.76 & 6.582 & 21.16 & 20.22 & 0.32 & ma97 127 v5188 & 11.3438 & 41.7022 & 1.52 & 6.776 & 21.31 & 20.38 & 0.23 & ma97 111 v1416 & 11.2661 & 41.6983 & 2.98 & 6.925 & 20.64 & 19.57 & 0.28 & v3568 & 11.3117 & 41.7760 & 2.10 & 7.165 & 21.31 & 20.73 & 0.35 & v5968 & 11.3586 & 41.7275 & 1.64 & 8.519 & 21.12 & 20.05 & 0.24 & ma97 114 v2242 & 11.2854 & 41.7508 & 1.12 & 8.680 & 21.47 & 20.51 & 0.26 & v7523 & 11.3917 & 41.6581 & 1.82 & 8.709 & 21.05 & 20.43 & 0.24 & ma97 121 v2276 & 11.2886 & 41.6656 & 1.42 & 9.803 & 20.86 & 19.70 & 0.17 & v7553 d31b v1791 & 11.2763 & 41.6948 & 3.04 & 10.011 & 20.35 & 19.58 & 0.26 & v6363 & 11.3683 & 41.6595 & 1.18 & 10.593 & 21.41 & 20.01 & 0.26 & ma97 117 v4733 & 11.3324 & 41.7888 & 1.99 & 10.971 & 21.39 & 20.11 & 0.35 & v9029 & 11.4211 & 41.7472 & 2.95 & 11.668 & 20.84 & 19.71 & 0.32 & ma97 126 v107 & 11.2255 & 41.7051 & 4.41 & 12.525 & 20.59 & 19.66 & 0.37 & v4104 & 11.3261 & 41.6504 & 1.45 & 12.557 & 21.21 & 19.98 & 0.23 & v3407 & 11.3085 & 41.7645 & 4.39 & 12.801 & 20.27 & 19.40 & 0.39 & v8530 & 11.4076 & 41.8088 & 0.79 & 12.809 & 21.38 & 20.17 & 0.22 & v8882 & 11.4152 & 41.8081 & 1.51 & 13.097 & 21.21 & 19.76 & 0.35 & v4407 & 11.3308 & 41.6660 & 1.70 & 14.112 & 19.97 & 19.14 & 0.14 & v6759 & 11.3772 & 41.6576 & 6.45 & 15.479 & 20.46 & 19.41 & 0.51 & ma97 118 v5760 & 11.3544 & 41.7088 & 4.49 & 16.608 & 20.83 & 19.69 & 0.44 & v5614 & 11.3509 & 41.7351 & 5.43 & 20.18 & 20.46 & 19.37 & 0.38 & ma97 113 v4452 & 11.3319 & 41.6506 & 1.37 & 26.59 & 21.56 & 19.52 & 0.38 & v6165 & 11.3632 & 41.6997 & 5.58 & 28.76 & 20.05 & 18.84 & 0.43 & ma97 116 v4711 & 11.3361 & 41.6596 & 1.15 & 32.29 & 21.54 & 20.06 & 0.38 & v5415 & 11.3483 & 41.6933 & 4.46 & 37.06 & 20.64 & 19.02 & 0.48 & ma97 112 v9679 & 11.4405 & 41.7749 & 2.46 & 42.55 & 20.18 & 18.75 & 0.33 & ma97 129 [ table : ceph ] in table [ table : misc ] we present the parameters of 11 miscellaneous variables in the m31a field , sorted by the decreasing value of the mean magnitude @xmath22 . for each variable
we present its name , 2000.0 coordinates , value of the variability index @xmath57 , mean magnitudes @xmath22 and @xmath23 . to quantify the amplitude of the variability
, we also give the standard deviations of the measurements in @xmath2 and @xmath10 bands , @xmath55 and @xmath56 . in the `` comments ''
column we give a rather broad sub - classification of the variability : lp possible long - period variable ( @xmath58 ) ; irr irregular variable . unlike the m31b field ( paper i ) ,
all of the miscellaneous variables seem to represent the lp type of variability .
lllllllllc v1494 & 11.2659 & 41.7755 & 1.30 & 23.15 & 21.56 & & 0.33 & & v410 & 11.2355 & 41.7199 & 1.10 & 28.62 & 20.92 & & 0.13 & & v1911 & 11.2800 & 41.6786 & 1.68 & 36.74 & 19.99 & 19.25 & 0.12 & 0.14 & rv tau v3368 & 11.3104 & 41.6779 & 2.62 & 46.18 & 21.14 & 20.34 & 0.36 & 0.23 & rv tau v2977 & 11.3018 & 41.6835 & 1.27 & 53.3 & 21.21 & 20.13 & 0.18 & 0.28 & rv tau v9659 & 11.4421 & 41.6874 & 1.75 & 56.6 & 20.77 & 19.96 & 0.40 & 0.27 & rv tau
[ table : per ] llllllllc v3222 & 11.3081 & 41.6494 & 1.40 & 19.11 & 17.21 & 0.09 & 0.04 & v8123 d31b v3901 & 11.3216 & 41.6633 & 1.61 & 19.25 & 16.72 & 0.10 & 0.05 & lp v2109 & 11.2844 & 41.6975 & 1.67 & 20.07 & 19.13 & 0.11 & 0.07 & lp v7718 & 11.3926 & 41.7313 & 2.02 & 20.23 & 19.59 & 0.15 & 0.10 & lp v8415 & 11.4050 & 41.7867 & 1.98 & 20.37 & 19.65 & 0.17 & 0.12 & lp v5764 & 11.3563 & 41.6507 & 1.36 & 20.75 & 20.20 & 0.14 & 0.12 & lp v2570 & 11.2912 & 41.7629 & 1.82 & 21.42 & 19.60 & 0.53 & 0.20 & lp v541 & 11.2424 & 41.6594 & 2.47 & 21.51 & 19.20 & 0.54 & 0.13 & v5897 d31b v11 & 11.2240 & 41.6504 & 1.27 & 21.72 & 19.31 & 0.47 & 0.12 & v5075 d31b v9459 & 11.4322 & 41.7689 & 1.71 & 21.84 & 19.23 & 0.54 & 0.18 & lp v1032 & 11.2565 & 41.6655 & 1.72 & 21.88 & 19.43 & 0.43 & 0.14 & lp [ table : misc ] the area of m31a field has not been observed frequently before and the only overlapping variable star catalog is given by magnier et al .
( 1997 , hereafter ma97 ) .
of 17 variable stars in ma97 which are in our m31a field , we cross - identified 16 . of these 16 stars ,
one ( ma97 120 ) we did not classify as variable ( @xmath59 ) and one ( ma97 123 ) was initially classified as variable ( @xmath60 ) , but then failed to classify as a cepheid .
the remaining 14 stars we classified as cepheids ( see table [ table : ceph ] for cross - identifications ) .
there was also by design a slight overlap between the m31a and m31b fields ( fig.[fig : xy ] ) .
out of two eclipsing binaries found in the overlap from the m31a field , v1555 d31a was cross - identified as v6913 d31b , with very similar properties of the light curves ( see table [ table : ecl ] ) .
v538 d31a turned out to have @xmath61 in the m31b database , i.e. it very narrowly escaped classification as a candidate variable star in this field .
there was only one cepheid from the m31a field in the overlap region , v2276 d31a ( also ma97 108 ) , and it was cross - identified as v7553 d31b , again with very similar properties of the light curves ( see table [ table : ceph ] ) .
there was a second cepheid in the overlap found in paper i , v7845 d31b ( @xmath62 ) , which failed to qualify as a cepheid in the m31a field because of the more stringent requirements for the reduced @xmath37 of the fit ( section 5 ) .
we also cross - identified three miscellaneous variables ( see table [ table : misc ] ) , out of four detected in the m31a field and five detected in the m31b field , which fell into the overlap region .
in fig.[fig : cmd ] we show @xmath63 color - magnitude diagrams for the variable stars found in the field m31a .
the eclipsing binaries and cepheids are plotted in the left panel and the other periodic variables and miscellaneous variables are plotted in the right panel . as expected , some of the eclipsing binaries occupy the blue upper main sequence of m31 stars , but there is a group of eclipsing binaries with @xmath64 .
the cepheid variables group near @xmath64 , with the exception of possibly highly reddened system v4452 d31a .
the other periodic variable stars have positions on the cmd similar to the cepheids .
the miscellaneous variables are scattered throughout the cmd and might represent few classes of variability , but most of them are red with @xmath65 , and are probably mira variables . in fig.[fig
: xy ] we plot the location of eclipsing binaries and cepheids in the fields m31a and m31b , along with the blue stars ( @xmath66 ) selected from the photometric survey of m31 by magnier et al .
( 1992 ) and haiman et al .
the sizes of the circles representing the cepheids variables are proportional to the logarithm of their period .
as could have been expected , both types of variables group along the spiral arms , as they represent relatively young populations of stars .
however , in the field m31a there is a group of cepheids located outside the population of the blue stars ( at @xmath67 and below ) .
also , these cepheids as well as the cepheids located in the inner spiral arm of m31 ( field m31b at ra@xmath68 ) have on average shorter periods than the cepheids located in the outer spiral arm
. we will explore various properties of our sample of cepheids in the future paper ( sasselov et al .
1998 , in preparation ) .
andersen , j. , 1991 , a&ar , 3 , 91 baade , w. , swope , h. h. , 1963 , aj , 68 , 435 baade , w. , swope , h. h. , 1965 , aj , 70 , 212 crotts , a. p. s. , & tomaney , a. b. , 1996 , apj , 473 , l87 freedman , w. l. , wilson , c. d. , & madore , b. f. , 1991 , apj , 372 , 455 freedman , w. l. , & madore , b. f. , 1990 , apj , 365 , 186 gaposchkin , s. , 1962 , aj , 67 , 334 haiman , z. , magnier , e. , lewin , w. h. g. , lester , r. r. , van paradijs , j. , hasinger , g. , pietsch , w. , supper , r. , & truemper , j. , 1994 , a&a , 286 , 725 hilditch , r. w. , 1996 , in : `` binaries in clusters '' , eds . e. f. milone & j .- c .
mermilliod ( asp conference series vol .
90 ) , 207 hubble , e. , 1926 , apj , 63 , 236 hubble , e. , 1929 , apj , 69 , 103 huterer , d. , sasselov , d. d. , schechter , p. l. , 1995 , aj , 100 , 2705 jacoby , g. h. , et al . , 1992 ,
pasp , 104 , 599 kaluzny , j. , stanek , k. z. , krockenberger , m. , sasselov , d. d. , tonry , j. l. , & mateo , m. , 1998 , aj , in press ( paper i , astro - ph/9703124 ) krockenberger , m. , sasselov , d. d. , & noyes , r. , 1997 , apj , 479 , 875 landolt , a. , 1992 , aj , 104 , 340 magnier , e. a. , lewin , w. h. g. , van paradijs , j. , hasinger , g. , jain , a. , pietsch , w. , & truemper , j. , 1992 , a&as , 96 , 37 magnier , e. a. , lewin , w. h. g. , van paradijs , j. , hasinger , g. , pietsch , w. , & trumper , j. , 1993 , a&a , 272 , 695 magnier , e. a. , augusteijn , t. , prins , s. , van paradijs , j. , & lewin , w. h. g. , 1997 , a&as , in press metzger , m. r. , tonry , j. l. , & luppino , g. a. , 1993 , asp conf .
ser . , 52 , 300 paczyski , b. , 1997 , in : `` the extragalactic distance scale stsci symposium '' , eds . m. livio , m. donahue & n. panagia ( cambridge : cambridge university press ) , p. 273 sasselov , d. d. , et al . , 1998 , in preparation stetson , p.b . 1987 , pasp , 99 , 191 stetson , p.b 1991 , in `` astrophysical data analysis software and systems i '' , asp conf . ser .
worrall , c. bimesderfer & j. barnes , 297 stetson , p. b. , 1996 , pasp , 108 , 851 tammann , g. a. , 1996 , pasp , 108 , 1083 tonry , j. l. , blakeslee , j. p. , ajhar , e. a. , & dressler , a. , 1997 , apj , 475 , 399 van den bergh , 1996 , pasp , 108 , 1091 | we undertook a long term project , direct , to obtain the direct distances to two important galaxies in the cosmological distance ladder m31 and m33 , using detached eclipsing binaries ( debs ) and cepheids .
while rare and difficult to detect , detached eclipsing binaries provide us with the potential to determine these distances with an accuracy better than 5% . the massive photometry obtained in order to detect debs provides us with the light curves for the cepheid variables .
these are essential to the parallel project to derive direct baade - wesselink distances to cepheids in m31 and m33 . for both cepheids and eclipsing binaries
the distance estimates will be free of any intermediate steps .
as a first step of the direct project , between september 1996 and january 1997 we have obtained 36 full nights on the michigan - dartmouth - mit ( mdm ) 1.3-meter telescope and 45 full / partial nights on the f. l. whipple observatory ( flwo ) 1.2-meter telescope to search for detached eclipsing binaries and new cepheids in the m31 and the m33 galaxies . in this paper , second in the series , we present the catalog of variable stars , most of them newly detected , found in the field m31a ( @xmath0 ) .
we have found 75 variable stars : 15 eclipsing binaries , 43 cepheids and 17 other periodic , possible long period or non - periodic variables .
the catalog of variables , as well as their photometry and finding charts , are available using the anonymous ftp service and the www .
the ccd frames are available on request . |
the description of the states that a pack of macroscopic objects can take has become an important subject of debate in physics since basic equilibrium statistical mechanics was proposed as a suitable framework to deal with this problem @xcite .
it is expected that macrostates consisting of a large collection of configurations ( or microstates ) compatible with given macroscopic variables that describe the macroscopic state can be readily defined and that the preferred macrostate for the system under given constraints would be predictable as the mean of a suitable distribution of such microstates .
the aforementioned collection of microstates have to be generated with a well defined protocol of repeated perturbations applied to the packing .
such perturbations are necessary since macroscopic objects ( such as the grains of a granular sample ) interact through non - conservative forces ( inelastic collisions and friction ) and so dissipate all kinetic energy at the particle scale .
hence , to move from one configuration to another , an external input of energy is mandatory .
whether a given protocol of perturbation leads to a collection of microstates that can be defined as an equilibrium macrostate in the sense described by thermodynamics is still a matter of debate .
nonetheless , there exist some consensus that in some cases this may be the case . for instance , annealed tapping protocols are known to give reproducible volumes ( and its fluctuations ) for a given tap amplitude and tap duration @xcite .
however , the question remains as to what are the macroscopic variables that fully describe the macroscopic state @xcite .
volume fluctuations present a controversial feature .
plotted as a function of volume , fluctuations display a maximum at an intermediate volume between the maximum and minimum volume reached by the macrostates in some studies @xcite .
this is in contrast with results from experiments using fluidized beds @xcite ( they show a minimum in fluctuations ) and experiments with non - jammed packings @xcite ( they present a monotonic decrease of fluctuations with increasing packing fraction ) . in this work
we exploit a model recently introduced by bowles and ashwin @xcite that allows for a full analytic calculation of the density of states of a frictionless granular packing .
the model has the advantage of corresponding to a realistic representation of an experimentally realizable system and then allows for a validation of the ability of the theory of ensembles to describe steady states attained by real granular samples . under edwards proposal , in the volume ensemble , all possible configurations compatible with a given macroscopic volume are equally probable .
however , the configurations sampled in the lab by using external perturbations can not be set at a prescribed volume .
therefore , it is generally assumed that the collection of states generated by a repeated perturbation , after any transient has fade , should be compatible with a `` canonical ensemble '' of constant compactivity ( the analogue of temperature ) .
we generate such collections of microstates by tapping at different intensities using realistic simulations of the bowles ashwin model .
then , we compare the results with the prediction of the `` canonical ensemble ''
. we will show that the probability distribution of microstates generated by tapping agree with the prediction of the `` canonical ensemble '' only for intermediate tap intensities . for high and low tap intensities ,
the simulation explores only a portion of the phase space available . despite the shortcomings of the ensemble theory , this solvable model predicts a non - trivial maximum in fluctuations that can be observed in the simulations but only for the frictional case
as a function of packing fraction @xmath0 for the bowles - ashwin model for @xmath1 ( black ) , @xmath2 ( red ) , @xmath3 ( blue ) and @xmath4 ( green ) .
( b ) the compactivity @xmath5 .
( c ) the volume fluctuations @xmath6 .
the black solid symbols indicate the position of the minimum ( up triangles ) and maximum ( down triangles ) @xmath0 allowed for each @xmath7 .
the corresponding packing fractions @xmath8 where entropy is maximum are indicated by arrows.,title="fig : " ] as a function of packing fraction @xmath0 for the bowles - ashwin model for @xmath1 ( black ) , @xmath2 ( red ) , @xmath3 ( blue ) and @xmath4 ( green ) .
( b ) the compactivity @xmath5 .
( c ) the volume fluctuations @xmath6 .
the black solid symbols indicate the position of the minimum ( up triangles ) and maximum ( down triangles ) @xmath0 allowed for each @xmath7 .
the corresponding packing fractions @xmath8 where entropy is maximum are indicated by arrows.,title="fig : " ] as a function of packing fraction @xmath0 for the bowles - ashwin model for @xmath1 ( black ) , @xmath2 ( red ) , @xmath3 ( blue ) and @xmath4 ( green ) .
( b ) the compactivity @xmath5 .
( c ) the volume fluctuations @xmath6 .
the black solid symbols indicate the position of the minimum ( up triangles ) and maximum ( down triangles ) @xmath0 allowed for each @xmath7 .
the corresponding packing fractions @xmath8 where entropy is maximum are indicated by arrows.,title="fig : " ] we consider a model granular system introduced by bowles and ashwin which has an analytic density of states @xcite .
a set of frictionless disk of diameter @xmath9 are arranged in a narrow channel of width @xmath10 .
a configuration mechanically stable against external pressure is obtained if the following local constraints are met .
first , every disk needs three contacts , one with one of the walls of the channel and two with the only two possible neighbors ( the possible branch vectors for any disk are either aligned against a wall or across the channel ) .
second , the two branch vectors for a given disk can not be both aligned against a wall ( see fig . [ model ] ) .
this allows to count all possible branch vector configurations ( each has a well defined volume ) using a simple binomial expression @xcite . as a result
, the entropy @xmath11 defined as the logarithm of the number of microstates associated with a given volume @xmath12 and number of disks @xmath13 is given by @xcite .
@xmath14 where @xmath15 , being @xmath16 the number of branch vectors aligned with the walls , and @xmath17 plays teh role of the boltzmann constant .
the volume @xmath12 occupied by the disks for a given @xmath18 is @xmath19 $ ] .
notice that @xmath12 can only take discrete values since @xmath18 is a rational with @xmath20 .
the compactivity @xmath5 is defined as the intensive variable conjugate to the volume , i.e. , @xmath21 , hence @xmath22}{\lambda [ 2\ln(1 - 2\theta)-\ln(1-\theta)-\ln(\theta)]}.\ ] ] the volume fluctuations characterized by the variance @xmath6 of the volume can be obtained from the `` specific heat '' as @xmath23 . therefore , @xmath24 ^ 2}{\left[\frac{4}{1 - 2\theta}-\frac{1}{1-\theta}+\frac{1}{\theta } \right]}.\ ] ] in fig .
[ entropy ] we show the entropy @xmath25 , the compactivity @xmath5 and the volume fluctuations @xmath6 as a function of the packing fraction @xmath26 for various values of @xmath7 .
notice that the range of allowed @xmath0 varies with @xmath7 .
the entropy presents a maximum at @xmath8 in agreement with other models for static packings @xcite and as estimated via simulations @xcite which corresponds to @xmath27 .
states for packing fractions below @xmath8 correspond to negative compactivities associated to the inversion population of these volume bounded systems .
some authors suggest these negative @xmath5 macrostates may be inaccessible , however , this does not need to be the case ; some preparation protocols may indeed lead to very low packing fractions @xcite .
interestingly , fluctuations present a maximum at packing fractions above @xmath8 .
this maximum fluctuation should be then observed even if only positive @xmath5 states are assessed .
notice , that the bowles
ashwin model is similar to the very first model proposed by edwards @xcite .
however , in edwards simplistic model , particles are assigned one of two possible local volumes without restriction associated with the volumes already assigned to the neighbors as it is done here .
we carry out discrete element simulations of disks of diameter @xmath9 subjected to the gravity force @xmath28 and confined in a narrow container of width @xmath7 .
the container has a flat base and is infinitely high .
the particle particle and particle wall interactions correspond to a normal restitution coefficient @xmath29 and a static and dynamic friction coefficient @xmath30 for the frictional simulations and @xmath31 for the `` frictionless '' simulations . in order to compare results from systems of different widths we vary the number @xmath13 of particles to ensure that a similar height of the granular column is obtained in all cases ( @xmath32 ) .
simultaneously , we tune the material density of the disks in such way that all systems have the same total mass .
we used the box2d library to solve the newton - euler equations of motion @xcite .
box2d uses a constraint solver to handle hard bodies . at each time step of the dynamics a series of 25 iterations
are used to resolve penetrations between bodies through a lagrange multiplier scheme @xcite . after resolving penetrations , the inelastic collision at each contact
is solved and new linear and angular velocities are assigned to the particles .
the equations of motion are integrated through a symplectic euler algorithm .
the time step @xmath33 used is @xmath34 @xmath35 .
solid friction is also handled by means of a lagrange multiplier scheme that implements the coulomb criterion .
previous works using this library have shown that simulations are consistent with other more complex interaction models for granular particles @xcite .
tapping is simulated by setting the initial velocity @xmath36 of the container ( originally at rest ) to a given positive value in the vertical direction . in doing so , the container , and the particles inside , moves upward and fall back on top of a zero restitution base . while the box dissipate all its kinetic energy on contacting the base , particles inside the box bounce against the box walls and floor until they fully settle . after all particles come to rest a new tap
is applied .
the intensity of the taps is characterized by the initial velocity imposed to the confining box at each tap ( i.e. @xmath37 ) .
the tapping protocol consists of a series of 46800 taps .
every 130 taps we change the value of @xmath38 by a small amount @xmath39 .
we initially decrease @xmath38 from about @xmath40 down to a very low value and then increase it back to its initial high value .
we repeat this protocol on the sample three times . at each value of @xmath38
the last 30 taps are used to average the packing fraction .
in fig . [ phi ] , we show the mean packing fraction @xmath0 as a function of the tap intensity @xmath38 for various system widths @xmath7 for frictional and frictionless disks .
the annealing protocol in which we start from high tap intensities yields a well defined mean packing fraction for any given @xmath38 .
hence , we have averaged results from all three repetitions of the annealing .
figure [ phi ] shows results consistent with previous studies where a minimum packing fraction has been reported for wider 2d systems and 3d systems @xcite .
the horizontal lines in fig . [ phi ] indicate the maximum and minimum values of @xmath0 allowed in the bowles ashwin model .
as we can see , tapping makes @xmath0 vary in a range much narrower than the possible theoretical values , both for frictional and frictionless disks .
frictional disks achieve higher and also lower packing fractions than frictionless disks .
interestingly , for @xmath1 , frictional disks reach mean packing fractions below the one predicted to correspond to the maximum entropy @xmath8 indicating that negative @xmath5 may have been achieved ( although we do not know the exact density of states for the frictional case ) . in all other cases ,
mean packing fractions remain above @xmath8 . to compare results from different @xmath7 , we plot in fig .
[ scaled - phi ] the same data as in fig .
[ phi ] scaled so that the minimum and maximum packing fraction in each curve corresponds to 0 and 1 , respectively [ i.e. , @xmath41 , and the position of the minimum @xmath0 also coincides ( i.e. , @xmath42 ) . as a function of scaled @xmath43 for frictional ( solid symbols ) and frictionless ( open symbols ) for @xmath1 ( black ) , @xmath2 ( red ) , @xmath3 ( blue ) and @xmath4 ( green ) . ]
the scaled @xmath44@xmath43 curve shows a good collapse in the full range of tap intensities considered .
this is an indication that , despite the large discrepancies in the range of densities achieve for different @xmath7 , tapping induces a similar exploration of configurations in all these systems . for frictional ( black ) and frictionless ( red ) disks .
( a ) @xmath45 , ( b ) @xmath46 , ( c ) @xmath47 .
the blue lines correspond to the analytical histogram @xmath48 $ ] where we have chosen @xmath49 to best fit the data for frictionless disks ( @xmath50 , @xmath51 , @xmath52 . ) .
the blue squares indicate the discrete packing fractions ( or volumes ) allowed by the bowles ashwin model for this 24-particle system.,title="fig : " ] for frictional ( black ) and frictionless ( red ) disks .
( a ) @xmath45 , ( b ) @xmath46 , ( c ) @xmath47 .
the blue lines correspond to the analytical histogram @xmath48 $ ] where we have chosen @xmath49 to best fit the data for frictionless disks ( @xmath50 , @xmath51 , @xmath52 . ) .
the blue squares indicate the discrete packing fractions ( or volumes ) allowed by the bowles ashwin model for this 24-particle system.,title="fig : " ] for frictional ( black ) and frictionless ( red ) disks .
( a ) @xmath45 , ( b ) @xmath46 , ( c ) @xmath47 .
the blue lines correspond to the analytical histogram @xmath48 $ ] where we have chosen @xmath49 to best fit the data for frictionless disks ( @xmath50 , @xmath51 , @xmath52 . ) .
the blue squares indicate the discrete packing fractions ( or volumes ) allowed by the bowles ashwin model for this 24-particle system.,title="fig : " ] the bowles - ashwin model considers frictionless disks and hence the mechanically stable configurations are reduced to a very small set . in fig .
[ hist ] , we can see the distribution of volumes obtained in the simulations for @xmath53 at three different tap intensities . for these histograms
, @xmath54 taps in the steady state for the selected @xmath38 have been collected . for frictionless disks ,
the volume takes only the discrete values predicted by the model .
however , frictional disks are able to arrange themselves in configuration having intermediate volumes since a disk may be stable with only two contacts thanks to friction even without contacting a wall . despite this
, frictional disks still take volumes predicted for frictionless disks in a much prominent fashion .
since we know the density of states [ @xmath55 , the volume histograms in the `` canonical ensemble '' can be calculated as @xmath56 $ ] @xcite . the blue lines in fig .
[ hist ] represent this histogram with @xmath5 chosen as a fitting parameter .
as we can see , low tap intensities yield histograms which are far from being described by the analogue of the system being coupled to a `` volume bath '' of given @xmath5 . for higher tap intensities ,
the assumption that tapping takes the system to a macroscopic state compatible with a bath at fixed @xmath5 is fair .
however , notice that high tap intensities lead to low @xmath5 and the system is unable to sample the smallest possible volume ( highest packing fraction ) in our simulations which has a non - zero probability according to the theory [ see fig .
[ hist](c ) ] .
figure [ fluct ] shows the volume fluctuations @xmath6 as a function of packing fraction for various widths of the container .
we have increased the statistics in fig .
[ fluct ] by running 20 independent instances to estimate error bars in the @xmath38 region of interest . as we discussed in section 2 , the bowles
ashwin model predicts a maximum in the fluctuations for a packing fraction above @xmath8 .
it is clear form fig .
[ fluct ] that fluctuation present a maximum for the frictional disks .
in contrast , for @xmath57 , frictionless disks under tapping sample higher values of @xmath0 where a monotonic decrease in @xmath6 is predicted . for @xmath1 , frictionless disks do sample states at @xmath0 low enough for the maximum in fluctuations to be observed .
although subtle , it can be seen that fluctuations grow slightly up to @xmath58 and decay quickly beyond that value .
the maximum in the fluctuations has been observed previously in quasi-2d experiments of tapped granular columns @xcite .
however , contrasting results have been found in 2d and 3d packings @xcite where a monotonic decay with increasing packing fraction has been observed , whereas in 3d fluidized beds a minimum in the fluctuations was reported @xcite . while dimensionality and jamming state may be the reasons behind such strong qualitative discrepancies between experiments , it is clear that in the present system , both , the analytic ensemble theory and the dem simulation of tapping agree in that a maximum in fluctuations exists .
figure [ fluct ] makes apparent an additional feature of fluctuations in the case of frictional disks .
different fluctuations are observed even if the mean volume of the macrostate is the same [ see parts ( b ) and ( c ) ] .
this has been observed before in similar systems @xcite .
the implication is that the macrostates are not fully described by the volume and further macroscopic variables are needed . in refs . @xcite it was concluded that adding the force moment tensor as an extra extensive variable allowed the states to be fully described having same fluctuations if the same mean volume and mean force moment tensor where observed for two states obtained through different tap intensities and tap duration
we have tested the prediction of the edwards ensemble for the bowles ashwin model by running dem simulations of tapped columns of frictionless and frictional disks .
these simulations of a realistic protocol of generating steady states for a model with analytic solution is particular suited to validate the ensemble approach to describe the statistics of static packings .
one interesting finding is that fluctuations are predicted by the ensemble theory to display a maximum as a function of @xmath0 and this has been indeed observed in the simulations . beyond the overall similarity between the theoretical and simulation results , there are clear differences that suggest the ensemble theory is unable to capture some of the response of the realistic system .
in particular , the range of volumes explored through tapping is much narrower than the full range of allowed volumes .
this feature akin to ergodicity breaking @xcite .
tapping may condition the system to explore only a region of the phase space .
however , we find that independent instances of the tapping protocol do not lead to sampling a different portion of the phase space as expected in ergodicity breaking systems such as glasses
. this would imply that the flat measure in the volume ensemble proposed initially by edwards is inadequate at least for this system ( but possibly for most granular systems ) .
a revision may be necessary to consider that each type of perturbation would induce the system to sample the phase space in a different way .
l.a.p . acknowledges the aspen center for physics where parts of this work were done , and s. s. ashwin for valuable discussions .
50 s. f. edwards and r. b. s. oakeshott , physica a * 157 * , 1080 ( 1989 ) .
e. r. nowak , j. b. knight , m. l. povinelli , h. m. jeager and s. r. nagel , powder tech .
* 94 * , 79 ( 1997 ) .
ribire , p. richard , p. philippe , d. bideau and r. delannay , eur .
j. e * 22 * , 249 ( 2007 ) .
l. a. pugnaloni , i. snchez , p. a. gago , j. damas , i. zuriguel , d. maza , phys rev e * 82 * , 050301(r ) ( 2010 ) .
l. a. pugnaloni , j. damas , i. zuriguel , d. maza , papers in physics * 3 * , 030004 ( 2011 ) . m. schrter , d. i. goldman and h. l. swinney , phys . rev .
e * 71 * , 030301(r ) ( 2005 ) .
j. g. puckett , f. lechenault and k. e. daniels , phys .
e * 83 * , 041301 ( 2011 ) .
s. s. ashwin , r , k. bowles , phys . rev
. lett . * 102 * , 235701 ( 2009 ) .
r , k. bowles , s. s. ashwin , phys .
e * 83 * , 031302 ( 2011 ) .
s. s. ashwin , m. z. yamchi , r. k. bowles , phys .
* 110 * , 145701 ( 2013 ) .
j. j. brey , a. prados and b. sanchez - rey , physica a * 275 * , 310 ( 2000 ) .
s. mcnamara , p. richard , s. de richter , g. le caer , and r. delannay , phys .
e * 80 * , 031301 ( 2009 ) .
m. pica ciamarra and a. coniglio , phys .
* 101 * , 128001 ( 2008 ) .
l. a. pugnaloni , m. mizrahi , c. m. carlevaro , f. vericat , phys .
e * 78 * , 051305 ( 2008 ) .
box2d physics engine , www.box2d.org e. catto , iterative dynamics with temporal coherence ( 2005 ) , available at http://box2d.googlecode.com/?les/gdc2005 erincatto.zip ( retrived on october 2010 ) .
carlos m carlevaro , luis a pugnaloni , j. stat .
mech : theo .
p01007 ( 2011 ) .
m. sanchz , c. m. carlevaro , l. a. pugnaloni , j. vib .
control , in press .
p. a. gago , n. e. bueno and l. a. pugnaloni , gran .
matt . * 11 * , 365 ( 2009 ) .
f. paillusson , d. frenkel , phys .
e * 109 * , 208001 ( 2012 ) .
m. pica ciamarra , a. coniglio , m. nicodemi , phys .
lett . * 97 * , 158001 ( 2006 ) . c. briscoe ,
c. song , p. wang and h. a. makse , phys .
* 101 * , 188001 ( 2008 ) . | we simulate via a discrete element method the tapping of a narrow column of disk under gravity . for frictionless disks ,
this system has a simple analytical expression for the density of states in the edwards volume ensemble .
we compare the predictions of the ensemble at constant compactivity against the results for the steady states obtained in the simulations .
we show that the steady states can not be properly described since the microstates sampled are not in correspondence with the predicted distributions , suggesting that the postulates of flat measure and ergodicity are , either or both , invalid for this simple realization of a static granular system .
however , we show that certain qualitative features of the volume fluctuations difficult to predict from simple arguments are captured by the theory . |
topological insulators ( ti ) are a new class of materials , which were theoretically predicted to exist in different bi based materials @xcite .
angle - resolved photoemission spectroscopy ( arpes ) has been used to verify the topological surface states @xcite of such materials .
ti have , due to strong spin - orbit coupling , an insulating bulk energy gap and generate conducting topologically protected gapless electronic surface states , which are robust against disorder and magnetic impurities @xcite .
these new materials have very interesting properties from the basic physics point of view and for future applications , for example the observation of the quantum spin hall ( qsh ) effect @xcite or the realization of majorana fermions for the application in topological quantum computation @xcite .
the alloy bi@xmath0se@xmath1 was proposed @xcite as a three - dimensional ( 3d ) ti , and later it was experimentally confirmed by arpes @xcite that this system possesses a single dirac fermion at the surface .
diverse investigations of the transport properties in bi@xmath0se@xmath1 were done , e.g. resistance measurements as a function of the temperature verify the metallic surface states @xcite , the hall effect confirms that bi@xmath0se@xmath1 is a n - type carrier and the magnetoresistance shows weak anti - localization ( wal ) @xcite .
the wal effect is used as evidence that the electrons moving at the surface of the material are time reversal symmetry protected . in a ti
the wal effect is a result of the strong spin - orbit coupling , which , in the presence of defects , puts the backscattering at a minimum when a magnetic field or a magnetic impurity is not present . under the absence of localization ,
the surface of a ti remains metallic and is expected to exhibit perfect wal due to helical surface states @xcite .
therefore 2d wal was used to verify the topological nature in several cases @xcite , where the temperature dependence of the coherence length @xmath4 is used to prove that the wal originated from the 2d surface states ( @xmath5 for 2d wal and @xmath6 for 3d wal ) @xcite . in the last years tremendous efforts
were realized in order to produce this kind of materials as thin films , nanowires , nanocrystals @xcite etc . , using different techniques such as molecular beam epitaxy ( mbe ) @xcite , metal - organic chemical vapour deposition @xcite including mechanical exfoliation @xcite or chemical vapour transport @xcite .
nevertheless , the preparation of a large area of bi@xmath0se@xmath1 films and the possibility to obtain a variety of morphology by the combination with electron beam lithography techniques is still a challenge .
the route of the amorphous - crystalline phase is a powerful method already used to produce high quality thin films of quasicrystalline alcufe @xcite , alpdmn @xcite and alpdre @xcite samples , whose existence in the phase diagram does not allow variations of more than 2% of the components from the correct concentration . to produce the amorphous films we use a method which allows us to obtain centimeter - large homogeneous samples .
the composition of the target material shows a maximum aberration of 1% when compared with the source material .
this technique is called the sequential flash - evaporation @xcite ( sfe ) and can be easily combined with electron beam lithography to obtain samples in the desired shapes , thickness and can also be used to produce hetero structures , e.g. for the study of proximity effects .
the initial material used for the thin film preparation was commercial bi@xmath0se@xmath1 millimetre - large pieces from alfa aesar with a purity of 99.999% .
after milling the pieces to 300500 @xmath7 m in diameter , the crushed powder was introduced into the flash evaporation device .
the flash evaporation process was done at room temperature under a nominal pressure of @xmath8 @xmath9 mbar .
the films for the transport measurements and x - ray analysis were prepared onto si - substrates @xmath10 coated with an amorphous insulating 150 nm thick si@xmath1n@xmath11 film .
tem results , a ) structure factor @xmath12 ( the inset shows the tem picture ) , b ) pair correlation function @xmath13 .
the inset shows the reduced atomic distribution function , the linear fit at @xmath14 nm is used to obtain the atomic number density . ]
x - ray diffraction of the crystalline and amorphous thin films .
the obtained lattice parameters are in good agreement with the literature values ( shown in brackets ) . ] for transmission electron microscope ( tem ) experiments , additional nacl crystals substrates were placed into the chamber for deposition , which , after dilution in bi - distilled water , left the films floating on the water surface . thereafter these films were collected onto a commercial tem cu - grid .
atomic structure analysis of the initial material and the prepared thin films was measured by means of x - ray diffraction . for tem measurements we used the transmission electron microscope philips
cm 20 feg , to obtain the structure factor s(_k _ ) and the pair correlation function g@xmath15 , we have used a standard method described in @xcite .
the resistance measurements were done using the conventional four point method ; in this case the @xmath16 size films were structured by scratching using a micro - manipulator .
the contacts for low temperature measurements were made out of indium and for the high temperature measurements we used pd .
the ohmic contacts were verified by measuring the i - v curves .
first low temperature measurements of the amorphous state were done in a standard commercial closed cycle refrigerator system with a minimal temperature of 25 k. the electrical resistance was measured using a keithley 6517a electrometer due to the high resistance of the amorphous thin films .
low - temperature measurements between 2 k and 275 k were realized in a commercial cryostat ( oxford ) implemented with a temperature controller ( linear research lr-700 bridge ) which allows for a temperature stabilization of 5 mk over all temperature range .
the cryostat is also equipped with a @xmath17 tesla superconducting magnet , which allows to measure the mr and hall - effect .
low - noise resistance measurements have been performed using the ac technique ( linear research lr-700 bridge ) with ppm resolution . in order to determine the crystallization temperature we used the dc technique ( keithley 2182 with 2001 nanovoltmeter and keithley 6221 current source ) on the previously structured samples . during the measurements a magnetic field
was applied perpendicular to the sample and current , and its value was monitored by a hall sensor installed close to the sample .
the thickness and stoichiometry of the films were measured with rutherford backscattering spectroscopy ( rbs ) @xcite and particle induced x - ray emission ( pixe ) @xcite using a 2.0 mev proton beam .
there is also an increasing interest in the properties of ti materials when magnetic impurities are present .
therefore , the amount of magnetic impurities was quantified using pixe @xcite measurements as well .
this method provides a high sensitivity , for the samples studied here the typical minimum detection limit is @xmath18g@xmath19 g .
the analysis of the rbs and pixe spectra was performed using xrump @xcite geopixe ii @xcite .
the 2.0 mev protons used have a penetration depth of @xmath20 in bi@xmath0se@xmath1 and therefore pixe gives the impurity concentration of all the complete sample if its thickness is less than this value , as it is in our case .
raman spectra of the crystalline samples were obtained at room temperature and ambient pressure with a dilor xy 800 spectrometer at 514.53 nm wavelength ( green ) and a 2 @xmath7 m spot diameter .
the incident power was varied between 0.5 mw to 3 mw in order to check for possible sample damage or laser induced heating effects .
no damage or significant spectral change was observed in this range of incident power .
after the sfe preparation process at room temperature , the amorphous samples were obtained . the electron diffraction pattern of the amorphous sample is shown in the inset of fig . [ fig : skgr ] .
after analysis we obtained the atomic structure factor @xmath21 and the pair correlation function @xmath13 , both results are plotted in fig .
[ fig : skgr ] .
the structure results are comparable to known similar amorphous semiconductor systems @xcite , but specially to the structure of the liquid phase of bi@xmath0se@xmath1 investigated by usuki @xcite .
our results show similar features as the liquid phase , e.g. the position of the first and second peak in the s@xmath22 and g@xmath15 are almost at the same values .
the first peak at @xmath23 nm in g@xmath15 corresponds to the same value in the liquid phase and is close to the corresponding first nearest neighbor distance in the crystalline phase of bi@xmath0se@xmath1 @xcite ; the peak @xmath24 nm was identified in the liquid phase as consequence of the building of a short - range order related to the crystalline phase @xcite .
after integration , the first peak in g@xmath15 allows us to calculate the coordination number @xmath25 .
this value is less than that of the liquid state ( @xmath26 ) @xcite and is due to the amorphous phase , which builds more defined first neighbors in comparison to the liquid phase . from the slope of the reduced atomic distribution @xmath27 $ ] at small values of _ r _ ( dashed line in fig .
[ fig : skgr ] ) , we can calculate the atomic number density @xmath28 nm@xmath29 , which is comparable to the value of the liquid phase when calculated using a simple linear extrapolation of the density of bi and se in the liquid state . from the structure results and the high temperature stability of the amorphous phase ( discussed later ) , we can conclude , according to the well established interpretation @xcite , that the system is a stabilized electron - phase , which means the formation of a pseudogap at the fermi energy @xcite is responsible for the high stability .
consequently , the influence on the electronic transport yields a very high resistivity and semiconducting - like behavior as a function of the temperature .
after annealing in argon atmosphere to 425 k , the sample transforms into the crystalline phase . in fig .
[ fig : xray ] we present x - ray results of the amorphous and crystallized thin films . in case of the crystalline films we can recognize the peaks corresponding to the rhombohedral @xmath30 @xmath31 structure of the bi@xmath0se@xmath1 phase .
the calculated lattice parameters are in very good agreement with literature data @xcite within experimental precision . using the scherrer equation
we can determine the crystalline grain size , for that we used the most intense peak and the calculation gives a value of @xmath32 nm .
taking into account that the unit cell is composed of three layers stacked along the @xmath33-axis , each known as quintuple layer ( ql ) , and that a ql has a thickness of @xmath34 nm , our films contain grains which include 20 qls .
that is important because in order to have a dirac cone it is necessary to have at least 5 - 6 ql @xcite .
resistance vs. temperature of the amorphous sample after preparation at 320 k. the inset shows the fit and the two contributions mentioned in the text . ]
resistance measurements of the as - prepared amorphous films were also done , the results are shown in fig .
[ fig : tdepamorph ] . in our device
the highest measurable value of the resistance has been reached at 30 k. in the complete temperature range the resistance shows an thermally mechanism , but with two different regimes , which can be seen in two different slopes shown in the inset of fig .
[ fig : tdepamorph ] .
resistance vs. temperature from 2 k to 275 k of the crystalline bi@xmath0se@xmath1 sample .
the upper left inset shows the resistance when the sample transforms from amorphous to crystalline phase , the intersection of the dash lines gives the crystallization temperature @xmath35 .
the inset on the r.h.s . shows the resistance of the crystalline sample . ]
we have found that the temperature dependence of the resistance can be well described by two different variable - range hopping ( vhr ) conduction processes . the general equation for vrh
is given by @xmath36,\ ] ] where @xmath37 is a prefactor and @xmath38 is a characteristic temperature coefficient .
the value of the exponent @xmath39 depends on the nature of the hopping process .
we found that in our amorphous bi@xmath0se@xmath1 the value of @xmath40 describes very well the measured transport behavior .
this value of @xmath39 was derived by mott @xcite , and that is in the case of a constant density of states at the fermi level . according to mott
, @xmath38 is a coefficient which depends on the density of states @xmath41 at the fermi level in the form : @xmath42 where @xmath43 is the localization length and @xmath44 is the boltzmann constant . as we mentioned before
, it is evident that the resistance in our sample shows two different temperature regions , in each one the resistance can be described by the mott - vrh . in order to describe the total temperature dependence we used a simple circuit containing two resistors in parallel which result
is shown in fig .
[ fig : tdepamorph ] as a line .
magnetoresistance vs. applied magnetic field , measured at different constant temperatures . ] a deviation of the fit from the experimental data at temperatures @xmath45 k can be observed because the sample was prepared at ambient temperature and above this value the resistance starts to decrease irreversibly .
the contributions of the two resistances are evident in the inset . from the fit
we obtain the characteristic temperature @xmath38 at every region . for the low temperature region @xmath46 k and for the high temperature region
@xmath47 k. using these values we can estimate the localization length @xmath43 if the value of @xmath41 is known , but this is not yet available in the literature . assuming a value of the same order for similar amorphous systems , @xmath48
, we obtain at low temperatures @xmath49 nm and at high temperatures @xmath50 nm . these values correspond approximately to the first and second nearest neighbor distances in the pair correlation function obtained from the tem investigations . by means of the vhr analysis
it is difficult to conclude on the reason of the change in the localization length at specific temperature ranges which remains an open question .
low temperature magnetoresistance , the lines show the fits using the wal theory . ]
coherence length @xmath4(nm ) vs. temperature @xmath51 and @xmath52 ( inset ) obtained from the fits of the magnetoresistance .
the red lines in the figures are the fits and the expected theoretical values of @xmath4 and @xmath52 respectively . ]
the transition from amorphous to crystalline phase was monitored by the measurement of the resistance during annealing in a home made device inside an oven under ar - flow .
the results are shown in the insets of fig .
[ fig : r - t ] .
the crystallization occurs in a nearly step - like process and it is evident that the as - prepared film has a larger resistance in comparison to the crystalline one .
the value of the crystallization temperature @xmath53 was determined by intersection of the two lines , one before and another one during the crystallization process .
we obtain a @xmath53 of @xmath54 k , which is relatively high in comparison to amorphous metallic systems @xcite but is in the same range of amorphous - like materials such as alcufe @xcite and alpdmn @xcite , where a pseudogap at the fermi energy was observed by means of angular - resolved photoemission studies @xcite and tunneling spectroscopy @xcite . in fig .
[ fig : r - t ] we present the results of the temperature dependence of the resistance of our film in the crystalline phase . at temperatures between 15 k and 285
k the sample shows the typical metallic behavior already observed in this system . in the low temperature range @xmath55
k we observe an increase in the resistance with decreasing @xmath56 , which nearly follows a logarithmic temperature dependence , as observed in some metals @xcite when a small amount of magnetic impurities is present . one possible explanation for this behavior would be an apparent kondo effect @xcite . note that the resistance increases nearly logarithmic below 12 k. at @xmath57 k the slope changes which can be a result of the strong influence of the large magnetic impurity content to the spin - orbit interaction characteristic for this system .
as consequence of the resistance behavior at low temperature we did some pixe measurements . we have analyzed the initial bulk material and the deposited thin film before and after it was transformed from the amorphous phase into the crystalline one .
the results of the elements with the highest concentration are summarized in table [ table : pixe ] . with the exception of fe , the concentration of all other magnetic elements was below the detection limit .
.[table : pixe]results of pixe measurements . [ cols="<,<,<,<",options="header " , ] hall resistivity vs. applied field at different temperatures .
the inset on the l.h.s
. shows the calculated mobility , the inset on the r.h.s . the calculated carrier density . ]
according to the pixe results our preparation method has not produced considerable change of the initial stoichiometry .
this result confirms the advantages of preparing with the sfe method as before mentioned . according to pixe , the fe ( cr ) content already present in the initial material
has increased more than one order of magnitude after preparation .
this contamination can only be due to the milling process before sfe , because the amount in the amorphous and crystalline are quite similar , therefore the contamination in the oven can be ruled out . in the following
we show the results of the transport properties under the application of a magnetic field . in spite of the magnetic impurity content and the large polycrystallinity of the film , the results are quite similar to those reported in the literature for this ti . the normalized magnetoresistance ( mr )
is defined as @xmath58 and are plotted in fig .
[ fig : mr - t ] . in general
the mr is positive for all measured temperatures and at high fields and high temperatures ( @xmath59 t , @xmath60 k ) it is dominated by a parabola like dependence , due to the lorenz force contribution to the carriers scattering under influence of a perpendicular magnetic field @xcite .
raman spectra of the crystalline bi@xmath0se@xmath1 sample .
the dashed lines indicates the characteristic modes . ] at temperatures @xmath61 k and in the low - field regime ( @xmath62 t ) the mr increases sharply as the magnetic field increases , which is the typical behavior of the weak anti - localization ( wal ) effect @xcite .
when applying a magnetic field , the time - reversal symmetry breaks down with increasing field strength and the electrons begin to backscatter such that the resistance rises . in order to fit our results with the theory
, we define the magnetoconductance ( mc ) as @xmath63 , with @xmath64 .
the low field part ( @xmath62 t ) of the mc can be well fitted using the standard wal theory developed by hikami - larkin - nagaoka , given as : @xmath65,\end{aligned}\ ] ] where @xmath52 is a coefficient whose value was predicted to be 1/2 in the case of systems with strong spin - orbit interaction , i.e. in the pure wal regime , @xmath66 is the digamma function , @xmath67 is the dephasing magnetic field related to the phase coherence length @xmath4 according to @xmath68 . using the eq .
[ eq : wal ] , we fit the experimental results , see fig . [
fig : mr - t2 ] , and estimate the different parameters at each temperature .
the prefactor @xmath52 and the phase coherence length are shown in fig .
[ fig : koho ] . according to our fit results , at the low temperatures @xmath69 ,
@xmath52 varies around @xmath70 , which is close to the theoretical expected value .
the high temperature values are between 0.38 and 0.45 , this can be a consequence of the diminishing of the coherence length with rising temperature .
the coherence length varies from 300 nm @xmath71 to 100 nm @xmath72 , which is in concordance with the value reported for bi@xmath0se@xmath1 thin films prepared by molecular beam epitaxy ( mbe ) with a thickness of 20 ql @xcite .
according to our results , the temperature dependence of the coherence length is @xmath73 , with @xmath74 .
the exponent gives information about the dimensionality of the system . according to the theory , for a two - dimensional system , @xmath75 and for one dimension , @xmath76 .
our result is close to the exponent expected for a two - dimensional system @xcite . for @xmath52 and the @xmath4 , the observed behavior and values obtained
are consistent with that reported for thin films @xcite and microwires of bi@xmath0se@xmath1 @xcite .
the hall measurements were done at the same temperatures as the mr and are presented in fig .
[ fig : hall ] .
its sign indicates that the major carriers are electrons , i.e. n - type over the investigated temperature range . from the slope we can calculate the 3d carrier density ,
the results are plotted in the inset of fig .
[ fig : hall ] .
the value obtained @xmath77 is close to that reported in @xcite where the sample was grown without dopants but also to @xcite where fe was used as dopant . according to theoretical predictions ,
the presence of magnetic impurities in a ti should open a small gap in the surface states due to the breaking of time - reversal symmetry @xcite . to verify these predictions experimentally ,
films containing fe in unknown concentrations ( but with less than 2 % according to the authors @xcite ) were produced . for all doped samples the results show the typical behavior of the kondo effect in the temperature dependence of the resistance .
the results of the magnetoresistance show the typical wal behavior and only in samples with a high concentration of fe , a mix of weak localization ( wl ) and wal effects has been seen @xcite .
anomalous hall effect was observed in bi@xmath0se@xmath1 thin films when doped with cr , but only when the concentration is @xmath78 1.4 % @xcite .
our hall effect results do not show any sign of an anomalous contribution as a consequence of the fe and cr content .
similar results were obtained when bi@xmath0se@xmath1 was doped with fe with a concentration of less than 2 % @xcite .
we conclude that the amount of fe and cr present in our sample apparently does not affect the expected behavior @xcite of the investigated transport properties of the ti .
finally , in the crystalline state we have done raman scattering spectroscopy of our films . according to the stoichiometry , the primitive unit cell of bi@xmath0se@xmath1 contains
five atoms , therefore 15 degrees of freedom are allowed at the center of the brillouin zone ( @xmath79 ) , from which 3 are acoustic modes and the rest are optical phonons ( @xmath80 , @xmath81 , @xmath82 and @xmath83 ) . according to previous work @xcite , e@xmath84 and a@xmath85 modes can be distinguished . in our films
we were able to detect the high frequency e@xmath86 and a@xmath87 modes , at @xmath88 and @xmath89 , respectively , see fig .
[ fig : raman ] .
the measured raman shift in our samples is in agreement with the results from literature @xcite , indicating that our films have similar structural quality as others produced with more sophisticated methods .
ti bi@xmath0se@xmath1 polycrystalline thin films can be obtained by annealing starting from the amorphous phase .
the preparation was done using a simple and inexpensive preparation method , which can also be used to produce similar ti materials , giving the chance to study diverse properties before , during and after the disorder - order transition .
the material in the amorphous phase has a high resistivity and behaves like a semiconductor , which transforms in to the metallic crystalline phase at temperatures higher than room temperature .
tem microscopy , x - ray and raman measurements were used to verify that we have produced the bi@xmath0se@xmath1 films in the desired crystalline phase . using pixe , a method with high accuracy
, we have quantified the presence of magnetic impurities ( such as fe ) in the initial material and in the prepared thin films . nevertheless , despite the films being polycrystalline and that magnetic impurities are present ,
all measured electronic transport properties characterizing a ti like bi@xmath0se@xmath1 were measured and our results are similar to those reported in literature . the fact that our results are quantitatively comparable with data obtained for samples prepared by more sophisticated methods and of apparently much better quality , indicates that : either the surface states are extraordinarily protected and robust against disorder and magnetic impurities , or the results we are comparing are not the ones that one should take as characteristics of a good ti .
future systematic measurements as a function of disorder and magnetic impurities would consequently clarify the amount of disorder and magnetic impurities or both , which causes bi@xmath0se@xmath1 to become as a non- ti material .
58ifxundefined [ 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 * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1557/proc-643-k8.3 [ * * , ( ) ] @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | we present a new method to obtain topological insulator bi@xmath0se@xmath1 thin films with a centimeter large lateral length . to produce amorphous bi@xmath0se@xmath1 thin films
we have used a sequential flash - evaporation method at room temperature .
transmission electron microscopy has been used to verify that the prepared samples are in a pure amorphous state . during annealing the samples
transform into the rhombohedral bi@xmath0se@xmath1 crystalline strcuture which was confirmed using x - ray diffraction and raman spectroscopy .
resistance measurements of the amorphous films show the expected mott variable range hopping conduction process with a high specific resistance compared to the one obtained in the crystalline phase ( metallic behavior ) . we have measured the magnetoresistance ( mr ) and the hall effect ( he ) at different temperatures between 2 k and 275 k. at temperatures @xmath2 k and fields @xmath3 t we observe weak anti - localization in the mr ; the hall measurements confirm the n - type character of the samples .
all experimental results of our films are in quantitative agreement with results from samples prepared using more sophisticated methods . |
in the last two decades many experiments on direct detection of dm have been performed .
many of them are running today providing new results which have imposed upper limits for the interaction cross - section of these particles with the stable baryons : neutrons and protons .
they use a great variety of materials and detection techniques .
however the neutrons are still very difficult to discriminate along with a good rejection for gamma rays coming from the natural radioactivity .
the cosmogonic activation is in all these experiments an important source of intrinsic contamination . in general the sensitive medium of an ideal detector of wimps
should have the following desired properties : * to be composed of odd nuclei in order to have as well as the scalar interaction ( coherent ) the axial spin - spin interaction with the hypothetical particles * to present a high neutron capture cross - section to sign the neutron background * to be produced with a high purity , mainly free of radioactive isotopes * to present a compton interaction cross - section as low as possible in order to minimize the interaction with the natural gamma radioactivity background in addition , the detection device should allow to get a threshold energy as low as possible . in the superfluid phase
b at ultra low temperatures ( t @xmath0 100 @xmath1 k ) presents in the lancaster s configuration @xcite a low detection threshold ( @xmath21 kev ) in a ultra high purity state .
the other features are largely fulfilled by the nucleus itself . nevertheless to further enhance the background rejection we propose a matrix configuration in which the correlations among cells allow to improve the discrimination of background events .
this rejection has been confirmed and quantitatively estimated by a monte carlo simulation study described below .
the primary device consisted of a small copper cubic box ( v@xmath0 0.125 @xmath3 ) filled with .
it is immersed in a larger volume containing liquid and thin plates of copper nuclear - cooling refrigerant , see fig.1 .
two vibrating wires are placed inside the cell , forming a lancaster type bolometer @xcite .
a small hole on one of the box walls connects the box to the main volume , thus allowing the diffusion of the thermal excitations of the generated by the energy deposited in the bolometer by the interacting particle . + this high sensitivity device is used as follows : the incoming particle deposits an amount of energy in the cell , which is converted into quasiparticles of the superfluid state .
these are detected by their damping effect on the vibrating wire .
it must be pointed out that the size of the hole governs the relaxation time ( quasiparticles escape time ) and the q factor of the resonator governs the rising time .
the present device has a rather high q factor ( q @xmath4 ) , giving a rising time of the order of one second .
although the primary experiment @xcite was still rudimentary , it has allowed to detect signals down to a threshold of 1 kev .
many ideas are under study to improve the sensitivity of such a cell .
recently , the fabrication of micromechanical silicon resonators has been reported @xcite and the possibility to use such wires at ultra - low temperatures is under study .
the performance of a detector for direct dm search is based on the discrimination capability of the different types of background events .
the main background components are : thermal and fast neutrons , muons and gamma rays .
the aim of the simulation described below was to evaluate the total rejection of different configurations of matrices of cells taking into account the correlation among the cells and the energy loss in each cell .
as a typical wimp is expected to transfer up to 6 kev , it is necessary to evaluate the proportion of background events releasing less than 6 kev in the cell .
+ the elastic cross - section between a wimp and being small ( @xmath5 ) , a wimp event will be characterized by a single - cell event , with equal probability among all the cells of the matrix .
hence , the rejection against background events will be achieved by choosing events having the following characteristics : * only one cell fired ( single - cell event ) .
the quality parameter related to this selection will be referred to as correlation coefficient .
* energy released in this cell below 6 kev ( energy rejection ) . * the outermost cell layer is considered as a veto ( off - line )
, hence the fired cell has to be in the inner part of the matrix .
this will reject neutrons interacting elastically .
the events , generated in a direction perpendicular to the upper face , are 10 kev neutrons .
it can be noticed that most of neutrons of this energy are captured in the first layer . ]
a simulation has been done , in order to define an optimum design as a function of the number of cells and the volume of each cell .
then , the rejection capability has been estimated , for the preferred design , as well as the false wimp rate . +
the simulation has been done with a complete monte - carlo simulation using geant3.21 @xcite package and in particular the gcalor - micap(1.04/10 ) @xcite package for slow neutrons .
the simulated detector ( fig . [ 10kev ] ) consists of a cube containing a variable number of cubic cells , immersed in a large volume containing ( @xmath6=0.08 @xmath7 ) .
each cell is surrounded by a thin copper layer and it is separated from the others by a 2 @xmath8 gap ( filled with ) .
+ in contrast to most dm detectors , mache3 may be sensitive to rather low energy neutrons , and its response depends strongly on their kinetic energies . the total cross - section interaction for a neutron in ranges from @xmath9 barns , for low energy neutrons(e@xmath10 1 ev ) , down to @xmath11 barn for 1 mev neutrons .
the main processes are : elastic scattering which starts being predominant above 600 kev , and neutron capture : ( n , p ) , which is largely predominant for low energy neutrons ( @xmath12 kev ) : @xmath13 the energy released by the neutron capture is shared by the recoil ions : the tritium @xmath14h with kinetic energy 191 kev and the proton with kinetic energy 573 kev .
the range for these two particles is fairly short : typically 12 @xmath1 m for tritium and 67 @xmath1 m for proton ; consequently neutrons undergoing capture in are expected to produce 764 kev within the cell , thus being clearly separated from the expected @xmath15 signal ( e @xmath16 6kev ) .
the tritium produced by neutron capture will eventually decay with a half - life of 12 years by @xmath17-decay with an end - point electron spectrum at 18 kev .
it means that the number of neutrons capture per cell must be counted to estimate the contribution of this kind of events on the false @xmath15 rate .
+ the neutron capture cross - section decreases with increasing neutron kinetic energy , but on the other hand , the energy released in the cell by the elastic scattering is getting larger , thus diminishing the probability to leave less than 6 kev .
we present here , the main quality parameters for 1 mev neutrons ( fig .
[ coeff ] ) and for 2.6 mev @xmath18-rays ( fig .
[ gam2.6 ] ) .
this coefficients depend both on the matrix and cell sizes .
the full simulation for various kinetic energies has shown @xcite that a large cell ( 125 @xmath19 ) allows to obtain a large energy rejection ( r@xmath20 for 1 mev neutrons ) , and a large matrix ( 1000 cells or more ) allows to have a good correlation among the cells ( @xmath21 65% for 1 mev neutrons ) , thus rejecting efficiently @xmath18-rays and neutrons . for background rejection consideration the optimum configuration is a matrix of 1000 cells of 5 @xmath22 side ( more details may be found in @xcite ) .
the different curves correspond to different matrix sizes as indicated by the labels . ]
a detailed simulation has been done for the matrix of 1000 cells of 5 @xmath22 side representing a good compromise between rejection power , volume and feasibility . for neutrons , a large rejection events ( less than 6 kev in one non - peripheric cell ) .
] is achieved below 8 kev ( mainly thanks to veto and energy rejection ) and above 1 mev ( thanks to correlation and energy loss measurement ) .
as expected , 8 kev neutrons represent the worst rejection ( see fig .
[ rejec ] ) , since the capture process is less predominant , and the energy left by ( n , n ) interaction is always less than 6 kev .
+ for gamma rays , a high granularity detector provides an intrinsic rejection ranging between 10 and 1000 , depending on their kinetic energies .
this selection , may be improved by adding a discrimination between recoils and electrons under study .
different experimental approaches should be tested shortly .
a complete study of an inner and outer cryostat shielding is also needed , as well as an evaluation of natural radioactivity of materials .
+ each ) .
the different set of points correspond to @xmath18-rays ( squares ) and neutrons ( circles ) .
it must be pointed out that the evaluated rejection is for a `` naked matrix '' , i.e. without taking into account any lead or paraffin shielding or any separation between electron and ion recoils .
it represents the capability of the matrix to reject background events by means of energy loss measurements and correlation considerations . ] as neutrons recoiling off nuclei may easily simulate a @xmath15 event , it is crucial to evaluate the neutron - induced false event rate . for this purpose ,
a simulation of a paraffin neutron shielding has also been done , in order to evaluate the expected neutron spectrum through this shielding .
we have used the measured neutron spectrum @xcite in _ laboratoire souterrain de modane _ , between 2 and 6 mev , , will be highly suppressed by the 30 @xmath22 paraffin shielding ] with an integrated flux of @xmath23 .
+ using this flux and the expected rejection factor ( fig . [ rejec ] ) , we evaluated the false @xmath15 rate induced by neutron background , to be @xmath210.1 false event per day through the 1.5@xmath24 surface of the detector ( 1000 cells of 125 @xmath25 ) , which is much lower than the expected @xmath15 rate ( of the order of @xmath211 day@xmath26 in a detector of this size , see below and @xcite ) . a study of the muon background has also been done .
we estimated @xcite the @xmath1-induced false @xmath15 rate to be of the order of 0.01 day@xmath26m@xmath27 , which is more than two orders of magnitude below the expected @xmath15 rate .
the plot showing the accessible cross - section interaction for a rate of 0.1 events per day and per kg as a function of the mass of the wimp assuming a maxwellian distribution of the wimp s velocities in the galactic halo is shown on fig .
[ lsp ] for 10 kg of . - , in pb , for 10 kg of as a function of the neutralino mass .
the curve showing the predicted cross - section as a function of the neutralino mass for the special case of a pure photino is also shown ] however we have to verify if the available models allowing to estimate the interaction cross - section in the case of the best candidate for the lightest susy particle ( lsp ) as the neutralino give a value accessible by this plot . doing that for one of the special cases in which the neutralino is a pure photino we get an overlap shown on fig . [ lsp ] .
this special case is one example in which only the axial interaction channel is available to detect the elusive particle .
as a conclusion , it can been said that a large matrix ( @xmath21 1000 cells ) of large cells ( 125 @xmath25 ) is the preferred design for a multi - cell superfluid detector . by means of the correlation among the cells and the energy loss measurement , a high rejection
may be obtained for @xmath18-ray , neutron and muon background . for background rejection purpose ,
the main advantage of a superfluid detector is to present a high rejection against neutron background , mainly because of the high capture cross - section at low energy ; as neutrons interact _ a priori _ like @xmath15 , they are the ultimate background noise for dm detectors .
+ g. r. pickett , in _ proc . of the european workshop on low temperature devices for the detection of low energy neutrinos and dark matter ( 1988 annecy , france ) _ , edited by l. gonzales - mestres and d. perret - gallix ( ed .
frontieres ) . | mache3 ( matrix of cells of superfluid ) is a project of a new detector for direct dark matter ( dm ) search .
a cell of superfluid has been developed and the idea of using a large number of such cells in a high granularity detector is proposed.this contribution presents , after a brief description of the superfluid cell , the simulation of the response of different matrix configurations allowing to define an optimum design as a function of the number of cells and the volume of each cell . the exclusion plot and the predicted interaction cross - section for the neutralino as a photino are presented . |
the interference effects between the @xmath5 and @xmath6 decay amplitudes in the time - dependent asymmetries of @xmath7 decaying into @xmath8 and @xmath9 final states allow for the determination of @xmath10+@xmath11 .
nevertheless the extraction of @xmath10+@xmath11 depends upon the knowledge of the absolute value of the ratio ( @xmath12 ) of the amplitudes ( e.g. : @xmath13 ) and of the strong phase difference , which nave not been determined experimentally so far .
furthermore , the ratio @xmath12 is expected to be rather small , of the order of @xmath14 0.02 . in addition the extraction of 2@xmath0 from the two - body decays suffers from a eight - fold ambiguity .
one can estimate the ratio @xmath12 by deducing for instance the branching fraction @xmath15 from the measured branching fraction @xmath16 .
this approach is valid up to a theoretical uncertainty related to su(3 ) breaking effects and to the size of annihilation processes which do not contribute to @xmath17 final states . using all the experimental constraints @xcite,@xcite ,
a tentative extraction of 2@xmath0 has been made , attributing 100@xmath18 uncertainty to these assumptions , and getting @xcite : 2@xmath19 ( the @xmath20 in front of the central value refers to the fact that 2@xmath0 is determined up to a @xmath21 ambiguity ) .
another way of measuring @xmath22 is to perform a time dependent dalitz analysis of the three - body decay : , as proposed in @xcite .
this method allows in principle to overcome the disadvantages encountered in the determination of 2@xmath0 from @xmath8 or @xmath9 decays .
first of all , in the region of the dalitz plane where most of the interference takes place the value of the ratio @xmath12 is expected to be of the order of 0.4 since it involves @xmath5 and @xmath6 colour suppressed diagrams .
furthermore @xmath22 can be measured with only a two - fold ambiguity as explained in @xcite .
in addition , this method could be theoretically clean , because the strong amplitudes and phases can be , in principle , measured from data .
early studies can be found in @xcite . in this paper
we present a feasibility study of the analysis under realistic conditions . unlike to previous studies
, we have used a model for the dalitz structure of the decays that uses the latest experimental measurements .
we have also included realistic @xmath23 tagging performances and explored the possibility of fitting the strong amplitudes and phases of the most relevant interfering resonances by using both the tagged and untagged events .
finally we have studied the effect of background in the determination of @xmath22 .
the measurement of @xmath22 can be achieved through a study of the time - dependent evolution of decay sensitive to that phase .
this can only be performed on tagged events , that means events for which the original flavour of the reconstructed @xmath7 is known . in a @xmath7 factory
this is possible through the determination of the flavour of the other @xmath7 in the event since the two @xmath7 mesons are produced in a coherent @xmath24 state . thus if we determine that at a time @xmath25 we have a , then the other @xmath7 in the event at the same time @xmath25 is a . in practice ,
one of the @xmath7 ( the ) is reconstructed in a flavour tagging state .
the time - dependent decay rates taking into account cp violation for a ( @xmath26 ) or ( @xmath27 ) tagging meson are given by : @xmath28 where @xmath29 is the lifetime , @xmath30 is the mixing frequency and @xmath31 a normalisation factor .
the parameter @xmath32 is non - zero if there is mixing - induced cp violation , while a non - zero value for @xmath33 would indicate direct cp violation .
the final state @xmath1 can be reached through the diagrams shown in figure [ diagrams ] and considering the mixing .
the parameters @xmath32 and @xmath33 of eq .
[ eq : fplus ] will depend on the position in the @xmath7 dalitz plot .
indeed , a given bin in the dalitz plot gets contributions from amplitudes and strong phases of @xmath34 and @xmath35 transitions .
the model assumed for the decay parametrises the amplitude @xmath36 at each point @xmath37 of the dalitz plot as a sum of two - body decay matrix elements and a non - resonant term according to the following expression : @xmath38 where @xmath39 ( @xmath40 ) indicates the cabibbo allowed ( suppressed ) decay in each point @xmath37 of the dalitz plot .
each term of the sum is parametrised with an amplitude ( @xmath41 or @xmath42 ) and a phase ( @xmath43 or @xmath44 ) .
the factor @xmath45 gives the lorentz invariant expression for the matrix element of a resonance @xmath46 as a function of the position @xmath37 in the @xmath7 dalitz plot ; the functional dependence varies with the spin @xmath47 of the resonance according to the isobar model @xcite .
the total phase and amplitude are arbitrary .
we have chosen amplitude unity and phase zero for the mode @xmath48 decaying into @xmath49 .
+ four transitions have to be considered associated to probabilities of connecting a or initial state to a @xmath50 or @xmath51 final states .
they are indicated in table [ tab : sumamp ] .
.__summary of the amplitudes involved in the analysis .
the index @xmath37 refers to the position in the dalitz plot .
the third column is derived from the second one using cp transformation . _ _ [ cols="<,<,<",options="header " , ] from this study the first conclusion is that the present measurements @xcite,@xcite are not sufficient to fix the dalitz model and give a precise indication on the error which can be obtained on 2@xmath0 .
in fact , the precision on @xmath22 strongly depends on the variation of the branching fractions of neutral @xmath7 into @xmath52 states within the measured errors .
there is also some dependence on the presence or not of non resonant contributions and on the values of strong phases for the most relevant resonances .
we also show the impact of the variation of @xmath53 from 0.3 to 0.5 . the model using only a non resonant contribution for the @xmath54 component
is also given as a reference since it was taken as example in the feasibility study presented in @xcite .
the corresponding error on 2@xmath0 is a factor 3 - 6 better with respect the one obtained using a set of realistic decay models .
this model has not been further considered as it is contradiction with present data @xcite .
an important feature of this method is the possibility to have a model independent determination of @xmath22 by fitting the parameters of the decay model .
we consider a sample corresponding to 1 of collected statistics .
the reference error is obtained as previously by fitting only @xmath22 and by fixing all the other parameters .
we try to progressively release the other parameters which characterise the decay model .
first we relax the @xmath52 and the non - resonant @xmath55 components .
we managed to fit the three amplitudes and phases and the error on @xmath22 increases by 10@xmath18 .
so far we have used only tagged events since they are the only one carrying information on @xmath22 . nevertheless also untagged events could play an important role in fitting the decay model parameters . in practice
we can add the following term to the likelihood given in eq.[timelikeli ] : @xmath56 the same fit is repeated with untagged events included in the sample .
the error on @xmath22 decreases by about 10@xmath18 . in the following untagged events are used .
+ the following test is to leave free the strong phases of the @xmath52 and the non - resonant @xmath54 components . in this case
the error on @xmath22 increases by about 20@xmath18 .
an extra increase of 10@xmath18 is obtained if the amplitudes of those components are further left free to vary in the fit .
one can also try to estimate the effect of the presence of the background in the dalitz plot . for simplicity
we take a flat background over the dalitz plot , being aware that reality could be different .
we assume that the discrimination between signal and background events is performed using additional variables ( _ e .
g. _ the reconstructed @xmath7 mass ) and we fix the signal over background ratio in the cp violation fit .
we run our simulation by considering a signal over background ( s / b ) ratio of 30@xmath18 or 50@xmath18 .
the result is that the error on @xmath22 increases by 25@xmath18 and 50@xmath18 , respectively .
furthermore if s / b is equal to 50@xmath18 it seems difficult to perform a fit where all @xmath55 and @xmath54 @xmath52 and the non - resonant components are left free in the fit . + as by - product of this analysis , the dalitz model can be fitted from the data if the level of background is not too large .
the @xmath55 @xmath52 related amplitudes and phases can be fitted with precisions lying between 20 - 30@xmath18 . furthermore both the phases and the amplitudes for @xmath52 from @xmath54 processes are determined with errors which can go up to 50@xmath18 .
in addition the excited k resonnaces ( @xmath57 ) which concerns only @xmath55 processes can be precisely determined from data with relative uncertainties better than 25@xmath18 .
we have performed a full feasibility study for measuring @xmath22 with time dependent dalitz analysis using @xmath1 final states . for this study
a realistic decay model , based on the available experimental information has been elaborated .
we conclude that the error on @xmath22 strongly depends on the decay model and that the currently available experimental information is not enough for a precise estimate . by varying the different parameters in the currently allowed range ,
the error on @xmath22 can vary within a factor of about 2.5 .
only a complete dalitz analysis could tell us where we stand .
this is possible on data .
+ in fact we have shown that if the level of background is not too large ( @xmath58 ) the full decay model can be obtained by fitting the dalitz plot using both tagged and untagged events . all the amplitudes and strong phases can be fitted and , in case of the dalitz decay model given in table [ tab : resonances ] , the precision on @xmath22 could be of about 50@xmath18 with 500 and should not be limited by the systematical uncertainties .
this implies that the precision on 2@xmath59 could be reduced at few percent level in case this analysis is performed at a high luminosity b factory @xcite .
we would like to thanks anne - marie lutz , patrick roudeau and riccardo faccini for useful suggestions and for the careful reading of this paper .
babar collaboration , b. aubert _ et al .
_ , [ hep - ex/0507075 ] .
r. aleksan and t. c. petersen , econf * c0304052 * ( 2003 ) wg414 [ hep - ph/0307371 ] .
utfit collaboration , m. bona _ et al .
_ , [ hep - ph/0501199 ] .
+ utfit collaboration , m. bona _ et al .
_ , jhep 0507 ( 2005 ) 028 [ hep - ph/0509219 ] . + also available at the url http://www.utfit.org , where results are continuously updated .
cleo collaboration , s. kopp _ et al .
_ , phys .
d * 63 * , 092001 ( 2001 ) ; cleo collaboration , h. muramatsu _ et al .
_ , phys .
lett . * 89 * , 251802 ( 2002 ) ; erratum - ibid : * 90 * 059901 ( 2003 ) . b. aubert _ et al . _
[ babar collaboration ] , [ hep - ex/0412040 ] .
babar collaboration , b. aubert _ et al .
_ , phys .
* 95 * 171802 2005 .
babar collaboration , b. aubert _ et al .
_ , phys .
96 * 011803 2006 .
babar collaboration , b. aubert _ et al .
_ , _ phys .
_ * d68*:092001,2003 .
j. albert et al .
infn - ae 05 - 08 infn road map report .
+ see also @xmath60 study group : + http://www.slac.stanford.edu/bfroot/www/organisation/1036@xmath61study@xmath61group/ and realted documents | in this paper we present a feasibility study for measuring the combination of the unitarity triangle angle 2@xmath0 with a time dependent dalitz analysis in decays using @xmath1 final states following the method proposed in @xcite . for this study
we build a model for this decay using the latest experimental information and we investigate the possibility of fitting together with 2@xmath0 all the relevant strong amplitudes and phases of the decay model .
1.5 cm 1.0truecm * f. polci@xmath2 , m .- h . schune@xmath3 and a. stocchi@xmath3
* 0.3truecm * @xmath4 dipartimento di fisica , universit di roma `` la sapienza '' + and infn , sezione di roma , * + + * @xmath3 laboratoire de lacclrateur linaire , * + + |
for three decades optically bright qsos were our only beacons in the high redshift universe . soon after the discovery of qsos it became clear that their comoving space density increases dramatically up to redshift two to three .
larger systematic surveys confirmed early suggestions that this evolution is reversed at even higher redshift and established an epoch of optically bright qso activity with a peak around redshift @xmath4=2.5 ( schmidt , schneider & gunn 1994 ; warren , hewett & osmer 1994 ; shaver et al . 1996 ; mcmahon , irwin & hazard 1997 ) .
qso activity is widely believed to be the result of accretion onto super - massive black holes at the centre of galaxies ( e.g. rees 1984 and references therein ) and a number of authors have linked the change in qso activity to a corresponding change in fuel supply at the centre of the host galaxies ( cavaliere & szalay 1986 ; wandel 1991 ; small & blandford 1992 ; yi 1996 ) .
haehnelt & rees ( 1993 ; hr93 henceforth ; see also efstathiou & rees 1988 ) recognized that the epoch of qso activity coincides with the time when the first deep potential wells form in plausible variants of hierarchical cosmogonies .
the past few years have seen dramatic improvements in detecting ordinary galaxies and starbursts out to @xmath53 , transforming our knowledge of galaxy and star formation in the high redshift universe .
there are also far more extensive data on the demography of supermassive black holes in nearby galaxies , and on low level activity of agn both in the optical and the x - ray bands .
we discuss here some implications for the formation and evolution of active galactic nuclei , attempting to tie these lines of evidence together in a consistent model .
the paper is organized as follows . in section 2
we discuss the general framework used to relate the luminosity function of qsos and star - forming galaxies to the mass function of dark matter ( dm ) haloes .
section 3 focuses on constraints on the accretion history of supermassive black holes .
section 4 summarizes observational implications and section 5 briefly discusses the respective role of stars and qsos for the re - ionization history of the inter - galactic medium .
section 6 contains our conclusions .
it is widely believed that the matter distribution in the universe is dominated by dark matter and that the gravitational growth of density fluctuations created in the early universe is responsible for the formation of the observed structures .
the strongest constraints on the fluctuation amplitude of the dm density field on small scales comes from the observed space density of rich galaxy clusters which probe only slightly larger scales than we are interested in here ( white , frenk & efstathiou 1993 ) .
we therefore take the four variants of cold dark matter ( cdm ) models chosen by jenkins et al .
( 1997 ) to comply with observational constraints in the low redshift universe on galaxy and cluster scales as a representative sample of viable models ( for parameters see table 1 ) .
the press - schechter formalism gives simple but still reasonably accurate estimates of the space density of dm haloes . in fig .
1 we plot the mass function of dm haloes at @xmath4=3 for the four models . as expected
the mass function is essentially a power - law at small masses and falls off exponentially at the high mass end .
while at @xmath4=0 , the mass functions for the different models are , by design , very similar , at @xmath4=3 the characteristic turnover mass in the mass function differs considerably between the models . in the next two sections we will explore the link between star - forming galaxies and qsos at high redshift , assuming that both populations trace the mass function of dm haloes .
.the model parameter of the cdm variants explored : @xmath6 is the _ rms _ linear overdensity in spheres of radius @xmath7 and @xmath8 is the shape parameter as defined in the fitting formula for cdm - like spectra by efstathiou , bond & white ( 1992 ) . [ cols="<,<,<,<,<,<",options="header " , ] steidel and collaborators ( steidel & hamilton 1992 ; steidel , pettini & hamilton 1995 ; steidel et al . 1996 ; giavalisco , steidel & macchetto 1996 ) detected a population of actively star - forming galaxies at redshift 2.5 to 3.5 with a space density of about @xmath9 similar to that of present - day galaxies and star formation rates of order @xmath10 .
the optical observations carried out correspond to the rest - frame uv and probe the star formation rate rather than the integrated stellar light .
little is known at present about the masses of these objects and the relation of their mass to the rate of detected star formation .
as figure 1 demonstrates , the space density of these star - forming galaxies corresponds to those of haloes with masses of @xmath11 and virial velocities of @xmath12 ( see also baugh et al .
further evidence for masses of this order comes from the strong clustering of these galaxies ( steidel et al .
1997 , bagla 1997 , jing & suto 1997 , frenk et al .
1997 , peacock 1997 ) .
a reasonable fit to the luminosity function of these objects can be obtained assuming a linear relation between star formation rate and halo mass , i.e. a constant mass - to - light ratio . in figure 2a
we plot such a fit for the scdm model .
a weakly non - linear relation is also consistent with the data and would indeed be required to match the shallow slope of the luminosity function at the high luminosity end reported recently by bershady et al .
the fits obtained for the other cdm variants are of the same quality . note , however , that the observed @xmath13 widths of these galaxies , @xmath14 ( pettini et al .
1997 ) , might be in conflict with the virial velocities of @xmath15 quoted above .
the space density of optically selected qsos at @xmath4=3 with @xmath16 is smaller than that of the detected star - forming galaxies by a factor of a few hundred .
it is therefore less obvious how to match the mass function of dm haloes with the b - band luminosity function of qsos .
as demonstrated by hr93 the well - synchronized evolution of optically selected qsos can be linked to the hierarchical growth of dm haloes on a similar timescale if the duration @xmath17 of the optically bright phase is considerably shorter than the hubble time .
the comoving space density of dm haloes hosting qsos then exceeds that of observed qsos by a factor @xmath18 where we have assumed that one new generation of haloes forms per unit redshift . for small @xmath17
this comes more and more in line with the predicted space density of dm haloes and that of star - forming galaxies at high redshift .
however , hardly anything is known about the masses of the host objects of optically selected qsos and this still leaves considerable freedom in the exact choice of @xmath17 . following the approach of hr93 we estimate the formation rate of active black holes by taking the positive term of the time derivative of the halo mass function and a simple parameterization for the black hole formation efficiency .
it is further assumed that active black holes radiate with a light curve of the form , @xmath19 , where @xmath20 is the ratio of bolometric to eddington luminosity and @xmath21 is the fraction of the bolometric luminosity radiated in the b - band .
we obtain reasonable fits for a wide range of lifetimes and for all the cdm variants if we allow ourselves some freedom in the relation between halo mass and black hole mass .
there are , however , systematic trends : with increasing lifetime the black hole mass has to become a progressively more nonlinear function of the halo mass and the black hole formation efficiency has to decrease in order to match the luminosity function of qsos .
this is due to the fact that qsos are identified with rarer and more massive haloes with increasing lifetime , and these fall on successively steeper portions of the halo mass function . in figure 2b and
2c we plot two specific choices of parameters which we denote as model a and b hereafter . for fig .
2b a qso lifetime close to the salpeter timescale @xmath22 and a scaling of black hole mass with halo virial velocity as @xmath23 was assumed ( @xmath24 is the total efficiency for transforming accreted rest mass energy into radiation ) . a physical motivation for this particular dependence is discussed in section 3 .
2c shows the case of a linear relation between the halo and black hole mass advocated by haiman and loeb ( 1997,hl97 ) which requires a qso lifetime of less than @xmath25 much shorter than the salpeter time for usually assumed values of @xmath24 .
figure 2 is for the scdm model while figure 3 is for the other cdm variants ( model a only ) . in principle
@xmath26 could also depend on mass or other parameters and the degree of agreement which can be obtained with our simple assumptions seems gratifying .
the last few years have seen tremendous progress in establishing the existence of supermassive black holes in our own galaxy and nearby galaxies .
there are now a number of excellent cases ( including that of our own galaxy ) where observations strongly imply the presence of a relativistic potential well ( watson & wallin 1994 ; miyoshi et al .
1995 ; genzel et al .
furthermore , the number of mass estimates has increased to a level where it becomes possible to make statistical arguments .
magorrian et al .
( 1997 ; mag97 henceforth ) published a sample of about thirty estimates for the masses of the putative black holes in the bulges of nearby galaxies .
mag97 confirm previous claims of a strong correlation between bulge and black hole mass ( kormendy & richstone 1995 , but see ford et al .
a linear relation of the form , @xmath27 , was obtained by mag97 as a best fit . however , considering the large scatter a mildly non - linear relation would probably also be consistent with the data
. we would further like to note here that a linear relation between black hole to bulge mass does not necessarily imply a linear relation between black hole and halo mass and as we will argue later a non - linear relation might be more plausible .
fugukita , hogan & peebles ( 1997 ) estimate the total mass density in stellar bulges as @xmath28 and together with the above ratio of black hole to bulge mass we get , @xmath29 considering the complicated selection biases of the mag97 sample , the small sample size and possible systematic errors in the black hole mass estimates this number is still rather uncertain .
van der marel ( 1997 ) e.g. emphasizes the sensitivity of black hole mass estimates to the possible anisotropy of the stellar velocity distribution and argues that the mag97 mass estimates might be systematically too high . nevertheless , as pointed out by phinney ( 1997 ; see also faber et al .
1996 ) @xmath30 exceeds the mass density in black holes needed to explain the blue light of qsos purely by accretion onto super - massive black holes , @xmath31 by a factor of about ten unless the value of @xmath32 is smaller than usually assumed ( soltan 1982 , chokshi & turner 1992 ) . while a few years ago it seemed difficult to discover the total mass in black holes necessary to explain the blue light emitted by qsos at high redshift , black hole detections in nearby galaxies now suggest that accretion onto supermassive black holes may actually be rather inefficient in producing blue light
there are three options to explain the apparently large value of @xmath33 : ( i ) @xmath30 is strongly overestimated , or ( ii ) @xmath34 during the optically bright phase is smaller than previously assumed , or ( iii ) supermassive black holes do not gain most of their mass during the optically bright phase .
a plausible solution with @xmath34 significantly smaller than 0.01 is discussed in the next section . here
we will explore the last possibility somewhat further .
the typical mass of a black hole at the end of the optically bright phase of duration @xmath17 exceeds that accreted during this phase by a factor @xmath35 .
this factor should be larger than 1 and smaller than @xmath36 and therefore , @xmath37 the question when supermassive black holes gained most of their mass is therefore closely related to @xmath17 and @xmath20 . for bright quasars
, @xmath20 must be @xmath38 ; otherwise excessively massive individual black holes would be required .
furthermore , @xmath20 will always be smaller than unity even if the ratio of the accretion rate to that necessary to sustain the eddington luminosity , @xmath39 , greatly exceeds unity .
this is because a `` trapping surface '' develops at a radius proportional to @xmath39 , within which the radiation advects inwards rather than escapes . in consequence ,
the emission efficiency declines inversely with @xmath39 for @xmath40 ( begelman 1978 ) .
for @xmath41 the possible range for @xmath17 is , @xmath42 if @xmath17 is very short ( as in model b ) and @xmath34 is not significantly smaller than 0.01 it seems inevitable that supermassive black holes have acquired most of their mass _ before _ the optically bright phase . we would like to point out here that a value of @xmath43 as large or larger than we infer from mag97 is actually needed for small @xmath17 ( @xmath44 ) . for the remainder of this section
we assume that @xmath17 is of order the salpeter time ( as in model a ) .
the ratio of accreted mass to total mass at the end of the optically bright phase is then equal to @xmath45 . if @xmath46 during the optically bright phase ( and if @xmath34 is not significantly smaller than 0.01 ) then the corresponding gain in mass by a factor @xmath47 indicated by mag97 has to occur _ after _ the optically bright phase .
as the accretion should not be optically bright the most plausible options are advection dominated accretion flows ( adafs ) and dust - obscured accretion ( narayan & yi 1995 , fabian et al .
1997 and earlier references cited therein ) .
adafs require low accretion rate with @xmath48 where @xmath49 and @xmath50 is the shakura - sunyaev viscosity parameter .
there is therefore a maximum growth factor for the black hole mass density due to adafs @xmath51 and @xmath52^{0.5}$ ] would be required even if the accretion lasts all the way from @xmath4=3 to @xmath4=0 . if , however , @xmath53 ( and @xmath54)then the gain in mass by a factor @xmath33 has to occur _ before _ the optically bright phase as in the case of small @xmath17 .
for adafs this would require @xmath55 and is therefore hardly plausible .
in this case dust - obscured accretion would be the only viable option .
a clue to this question might lie in the fact that no black holes smaller than @xmath56 have been detected and that there might be a physically determined lower limit to the mass of a supermassive black hole .
a natural lower bound would be a few times @xmath57 the upper limit for which a supermassive star can ignite hydrogen before it undergoes the post - newtonian instability ( see bond , carr & arnett 1984 and references therein ) .
to see this imagine the following scenario .
the gas in a newly collapsing halo ( which has not yet acquired a central black hole ) settles into a self - gravitating disk .
the rotational velocity of the self - gravitating disk can be related to the virial velocity of the halo @xmath58 as follows , @xmath59 where @xmath60 is the usual angular momentum parameter characterizing the halo , @xmath61 the ratio of disc to halo mass and @xmath62 the ratio of disc to halo specific angular momentum ( see also mo , mao & white 1997 ) . for an isothermal dm halo @xmath63
is required for the disc to be self - gravitating ( for flatter central profiles , self - gravitation is even more likely ) .
the self - gravitating disc will be subject to violent gravitational instabilities ( shlosman , begelman & frank 1990 ) . for simplicity
let us assume that the rotation velocity is constant with radius .
the rate of accretion due to gravitational instabilities can then be written as @xmath64 which is also constant with radius and where we have defined @xmath65 as the ratio of dynamical to accretion timescale .
rather small values of @xmath66 are already sufficient for appreciable accretion rates if the rotational velocity of the disk is as large , or larger than , the virial velocity of the dm halo .
the first @xmath67 can be transported to the centre on a time scale shorter than @xmath68 ( the lifetime of hydrogen burning star radiating at the eddington limit ) if @xmath69 .
the supermassive star will then collapse due to the post - newtonian instability before it runs out of fuel .
the accretion rate onto the newly - formed black hole will exceed its eddington accretion rate at first by a factor @xmath70 . in this phase
the object will radiate at the eddington limit and the luminosity increases linearly .
the subsequent evolution will depend on whether the emitted energy can back - react on the accretion flow prior to fuel exhaustion .
such a back - reaction ought to occur when the luminosity exceeds the energy deposition rate necessary to unbind the disc on a dynamical timescale which scales as @xmath71 .
this will limit the black hole mass to , @xmath72 where @xmath73 is the fraction of the accretion luminosity which is deposited as kinetic energy into the accretion flow ( cf .
silk & rees 1998 ) .
the accretion rate will then drop dramatically .
the back - reaction timescale will be related to the dynamical timescale of the outer parts of the disk and/or the core of the dm halo and should set the duration of the optically bright phase .
it is interesting to note here that the accretion rate will change from super - eddington to sub - eddington without much gain in mass if the back - reaction timescale is shorter than the salpeter time .
the overall emission efficiency is than determined by the value of @xmath39 when the back - reaction sets in and is reduced by a factor @xmath74 compared to accretion at below the eddington rate .
this results in a low overall value of @xmath34 which is as we discussed in the previous section
one option to explain a large value of @xmath75 .
subsequently the accretion rate will fall below the critical rate permitting an adaf , and the black hole will spend most of its time accreting with low efficiency . as discussed above it will depend on the value of @xmath76 whether black holes can gain most of their mass during this late phase of accretion . as is apparent from equation ( 6 ) the black hole mass is very sensitive to the value of @xmath77 .
numerical simulations have demonstrated that the distribution of @xmath78 depends very little on any halo properties ( barnes & efstathiou 1987 ; lemson & kauffmann 1997 ) .
numerical simulations , however , also show that angular momentum is very efficiently transferred from the gas to the dm halo unless considerable energy and/or momentum input prevents the infalling gas from clumping ( navarro & steinmetz 1997 ) . a moderate change in the ability of the gas to transfer angular momentum to the dm halo and in the amount of gas settling into a self - gravitating disc with redshift would result in a strong redshift dependence of the black hole formation efficiency . such a strong redshift dependence was proposed by hr93 to explain the rapid decline of the typical qso luminosity at @xmath79 . as discussed by cavaliere & vittorini ( 1997 ) the decrease of the ability of the gas to cool once the hierarchical build - up of dm haloes has progressed from galaxy to group and cluster scales and the reduced merger rates of galaxies compared to that of dm haloes contribute to this rapid decline .
the redshift evolution of the black hole formation efficiency and the expected scatter in @xmath78 can easily explain the large scatter in the observed black hole to bulge mass relation .
furthermore , it seems likely that the bulge - to - disk ratio of a galaxy is also related to the value of @xmath77 .
this fits in well with the fact that the black hole mass seems to correlate better with the bulge mass than the total mass of the galaxy and naturally explains the correlation of black hole formation efficiency with the hubble type of the galaxy .
obviously the evolution described in this section need not to occur strictly in that time sequence .
in hierarchical cosmogonies , where dm haloes undergo continuous merging the process will be re - started several times .
at which step of the above scheme the evolution starts depends on whether the accretion rate in the core of a newly - formed dm halo on a given level of the hierarchy is super - eddington , sub - eddington or sub - critical with respect to the supermassive black hole already present .
a question which we have not addressed here but that certainly deserves more attention is the merging history of supermassive black holes .
here we would only like to remark that in models with a strictly linear relation of black hole to halo mass constant with time ( as in model b ) the growth of black holes will actually be dominated by the merging of black holes , while a non - linear relation ( as in model a ) leaves room for a substantial gain in mass by accretion of gas at each step of the hierarchy .
figure 4 sketches two possible accretion histories with a low overall efficiency for producing blue light .
the solid curves describe an accretion history where most of the mass is accreted during the adaf phase while the dashed curves are for an accretion history where the black hole gains most of its mass during a short - lived early phase with with @xmath80 .
the panels show ( from top to bottom ) mass accretion rate in terms of the eddington accretion rate , the mass relative to the final mass and the optical and hard x - ray luminosity .
the accretion rate is constant at the beginning with @xmath40 .
the mass is therefore linearly rising and @xmath39 decreases .
the spectral energy distribution for accretion with @xmath80 is rather uncertain ( as indicated by the three parallel lines for @xmath81 in the two bottom panels ) and should depend on the absorbing column and the dust content of the outer parts of the self - gravitating disc and/or the host galaxy .
the sharp drop of @xmath39 marks the onset of the back - reaction on the accretion flow and either the start or the peak of the optical bright phase ( with a rather inefficient production of hard x - rays ) .
once the accretion rate has fallen below the critical rate for an adaf ( indicated by the dashed lines in the top panel ) the spectral energy distribution will change to one peaked in the hard x - ray waveband .
in figure 5 we plot some properties of the host objects predicted by our fiducial models for the different cdm variants . as expected in model a , where @xmath17 is longer , the halo masses are considerably larger than in model b especially at low luminosities .
this is also reflected in the expected luminosity of the host galaxy .
while in model a qso hosts ( @xmath82 ) would sample only the bright end of the luminosity function of star - forming galaxies ( @xmath83 ) , in model b the faintest qsos would reside in objects well beyond the current spectroscopic limit at these redshifts ( fig .
large @xmath17 seem therefore to stand a better chance to be consistent with the large host galaxy luminosities of bright qsos reported by aretzaga , terlevich & boyle ( 1997 ) .
figure 3c illustrates the `` rareness '' of host haloes / galaxies in terms of the peak height in a gaussian distribution . in model
a qso hosts correspond to significantly rarer peaks .
the clustering strength of haloes depends mainly on the rareness of the peaks ( kaiser 1984 , mo & white 1996 , bagla 1997 ) and the two models should therefore differ strongly in the predicted clustering length of qsos . with the upcoming larger surveys such as the 2df and the sdss it should be possible strongly to constrain the clustering length , and thus also @xmath17 .
the strong clustering recently confirmed by la franca & cristiani ( 1997 ) seems already to dis - favour small @xmath17 .
as pointed out by many authors , the x - ray emission of optically selected qsos is too soft to explain the hard x - ray background .
di matteo & fabian ( 1996 ) and yi & boughn ( 1997 ) argued that the emission from adafs has a spectral shape similar to the hard x - ray background .
fabian et al . ( 1997 ) suggested that this might also be true for dust - obscured accretion .
it is therefore tempting to link the rather large value of @xmath84 inferred from mag97 to the origin of the hard x - ray background and the recently detected large space density of faint x - ray sources ( almaini et al .
1996 ; hasinger et al .
1997 , schmidt et .
al 1997 , mchardy et al .
1997 , hasinger 1998 ) .
the presence of extremely low - level optical agn activity in a large fraction of galaxies reported by ho , fillipenko & sargent ( 1997 ) would also fit in nicely with such a picture .
the present - day black hole mass density is sufficient to produce the hard x - ray background if @xmath85 . the efficiency of adafs is @xmath86 and decreases rapidly for small @xmath50 and small @xmath87 .
if the hard - xray background was produced by adafs onto supermassive black holes in ordinary galaxies this requires a value of @xmath30 as high as we infer from mag97 , a large value of @xmath50 and a value of @xmath39 below but still close to @xmath88 lasting for a hubble time for the majority of all supermassive black holes .
one should note here that at present it is not easy to discriminate observationally between dust - obscured accretion and an adaf as little work has been done on the detailed spectral shape of an adaf embedded in an agn environment .
furthermore , at faint flux levels and high redshifts a possible star - burst contribution to the total spectral energy distribution will become more and more important in the optical and probably also the soft x - ray .
in section 3.2 we speculated that there might exist a minimum black hole mass .
such a minimum mass will have a strong bearing on the number of point sources at faint magnitudes .
6 shows the predicted number of sources per unit redshift per arcmin@xmath89 above a certain flux limit .
we have assumed that the sources emit a fraction @xmath90=0.3 of the eddington luminosity in the corresponding band .
no correction for dust absorption or intervening ly@xmath50 absorption has been made . since the universe becomes optically thick to ly@xmath50 absorption at @xmath91 these faint sources
could only be detected longward of the redshifted ly@xmath50 wavelength ( i.e k band or redder ) . for model a a minimum black hole mass of @xmath92 and for both models
a minimum halo virial velocity @xmath93 was assumed . at magnitudes
@xmath94 model a predicts a progressively smaller number of faint sources than model b mainly due to the assumed minimum black hole mass .
as pointed out by hl97 the next generation space telescope ( ngst ) should be an excellent tool to probe the black hole mass spectrum at high redshifts and small masses .
the rather small number of very faint point sources in the hdf might already be in conflict with a linear relation of black hole to bulge mass which extends below @xmath92 and to high redshift ( madau , private communication ; see also almaini & fabian 1997 ) .
the contribution from newly - formed stars relative to that from super - massive black holes to the re - ionization history of the universe is still rather uncertain . at @xmath4=3 the flux of ionizing photons attributed to observed qsos
is probably just sufficient to explain the observed uv background at the hydrogen lyman limit if the ionizing flux is as low as indicated by recent studies of the flux decrement distribution in qso spectra which make use of hydrodynamical simulations ( rauch et al .
1997 ; bi & davidsen 1997 ) .
rauch et al .
, however , find that the ionizing background necessary to fit the mean flux decrement decreases more slowly towards higher redshift than the expected contribution of observed qsos .
this indicates that at @xmath95 the ionizing background might be dominated by stars .
direct estimates of the stellar contribution to the ionizing background are still difficult as we do not know how much ionizing flux is actually escaping star - forming galaxies at high redshift .
just longward of the hydrogen lyman limit the photon flux due to star - forming galaxies exceeds that due to quasars by a factor of about thirty .
the spectra of star - forming galaxies certainly have a strong intrinsic lyman break but it is still unclear whether the ionizing flux from galaxies exceeds that due to qsos even at @xmath96 . a more precise assessment of this question will has to await detailed spectra of star - forming galaxies which extend beyond the hydrogen lyman limit .
theoretical predictions for the relative contribution of stars and qsos to the uv background are also hampered by our lack of knowledge of the photon escape fraction in both star - forming galaxies and qsos , its possible dependence on halo / galaxy properties and its evolution with redshift . in figure 7 , we have plotted predictions for the photon generation rate just longward of the hydrogen lyman limit evolving our two fiducial models backwards in time and assuming a minimum black hole mass of @xmath92 ( units are photons per baryon per unit redshift ) . in model
a the black hole formation efficiency strongly decreases with halo virial velocity .
in such a model the ionizing background should eventually be dominated by stars at high redshift . in model b ,
the fraction of `` blue '' photons coming from by qsos is only weakly redshift dependent .
this is primarily due to the linear scaling of the black hole and halo mass .
as argued by hl97 , in such a model the ionizing background could be dominated by qsos at all redshifts .
model predictions for the re - ionization redshift are even more uncertain and as discussed by many authors they differ strongly between different cdm variants and depend strongly on the assumed escape fraction
. while mixed - dark matter models are just capable of re - ionizing hydrogen at or before redshift five , in open models hydrogen re - ionization by stellar radiation might occur well before redshift ten .
further clues to the question of how and when the universe was re - ionized might be obtained from the thermal history of the inter - galactic medium and the closely related issue of the re - ionization history of helium .
the spectrum of radiation emitted from star forming regions is generally believed to be too soft to fully ionize helium .
if hydrogen were re - ionized by stars , then helium would have to be re - ionized by the harder output from qsos which might have occurred considerably later .
recently , there has been some evidence for a change in the spectral shape of the ionizing background at redshift three and only partial re - ionization of helium at the same redshift has been reported ( songaila & cowie 1996 , reimers et al .
1997 , but see miralda - escud 1997 for a discussion and also boksenberg 1997 ) .
however , haehnelt & steinmetz ( 1997 ) use an investigation of the observed doppler parameter distribution of qso absorption lines to argue that helium was probably fully re - ionized by uv radiation with a qso - like spectrum before @xmath4=4 .
the optical qso luminosity function at @xmath97 can be plausibly matched with the luminosity function of star forming galaxies at the same redshift and the mass function of dm haloes predicted by a range of variants of cdm cosmogonies believed to comply with observational constraints in the low - redshift universe .
this is possible for lifetimes of optically bright qsos anywhere in the range @xmath0 to @xmath1 .
there is a correlation between the lifetime and the required degree of non - linearity in the relation between black hole and halo mass .
the non - linearity has to increase for increasing lifetime .
predicted host halo masses , host galaxy luminosities , and the clustering strength all increase with increasing lifetime and further observations of these offer our best hope of constraining the duration of the optically bright phase of qsos .
the present - day black hole mass density implied by the integrated luminosities of optically bright qso may be significantly smaller than that inferred from recent black hole estimates in nearby galaxies for generally assumed efficiencies for producing blue light .
we have discussed three possibilities for how and when this mass could be accreted in an optically inconspicuous way : ( i ) in the early stages of accretion at rates far above the eddington rate , ( ii ) by accretion where optical emission is obscured by dust , or ( iii ) in the late stages of accretion at a rate below the critical rate for an advection dominated accretion flow with an shakura - sunyaev parameter of @xmath98 .
for the latter two possibilities the space density of faint x - ray sources should exceed that of optically selected qsos considerably and might reach that of present - day @xmath99 galaxies at the faintest flux levels .
the expected shape of the luminosity function of faint x - ray sources and its evolution with redshift should , however , differ and improved determinations of this luminosity function together with broader spectral coverage of the sources should soon determine to what extend one or more of these possibilities are realized . we have speculated that the formation and accretion history of supermassive black holes is determined by accretion at the centre of a gravitationally unstable self - gravitating disc in the core of a newly - formed dark matter halo and could proceed in the following steps : * accumulation of enough material to form a supermassive star , * growth to the limit for post - newtonian instability and collapse to a supermassive black hole of @xmath92 on a timescale less than @xmath68 , * steady accretion at well above the eddington rate for about @xmath3 with linearly increasing mass , * reduction of the accretion rate due to back reaction on the accretion flow once the luminosity exceeds @xmath100 by a sufficient factor , and * late accretion for a hubble time at a ( slowly decreasing ) reduced level .
this accretion history can explain the rather short optically bright phase , can account for a low overall efficiency for producing blue light and leaves room for the production of the hard x - ray background .
it gives a physical explanation for the mass of the black hole formed and the predicted scaling with virial velocity / mass of the halo that is required to match the optical qso luminosity function with the mass function of dm haloes for a duration of the optical bright phase of @xmath3 .
the rapid redshift evolution of the black hole formation efficiency can be attributed to a modest evolution in the fraction of the gas settling into a self - gravitating configuration at the centre of a dm halo and the ability of the gas to transfer angular momentum to the halo .
it might further have the interesting implication of a minimum black hole mass of about @xmath92 .
the predicted decreasing black hole formation efficiency towards smaller halo masses and the presence of a minimum black hole mass may suggest that the ionizing background is progressively dominated by stars at high redshift .
we would like to thank george efstathiou , andy fabian , jordi miralda - escud and max pettini for helpful discussions and comments on the manuscript .
almaini o. , fabian a.c
. , 1997 , mnras , 288 , l19 almaini o. , boyle b.j . , shanks t. , griffiths r.e . , roche n. , stewart g.c . , geogantopoulos i. , mnras , 1996 , 282 , 295 aretzaga i. terlevich r.j .
boyle b.j . , 1997 , mnras , in press ,
astro - ph/9711130 bagla j. s. , 1997 , submitted to mnras , astro - ph/9711081 barnes j. , efstahthiou g. , 1987 , apj , 319 , 575 baugh c.m .
, frenk c.s .
, lacey c. , 1997 , submitted to apj , astro - ph/970311 begelman m. c. , 1978 , mnras , 184 , 53 bershady m. , majewski s.r . , koo d.c . , kron r.g . , munn a. , 1997 , apj , in press , astro - ph/9709216 bi h. , davidsen a.f . , 1997 , apj , 479 , 523 boksenberg a. , in petitjean p. , charlot s. , eds , structure and evolution of the igm , 13th iap colloquium , in press , astro - ph/9710258 bond j.r , arnett w.d . , carr b.j . , 1984 ,
apj , 280 , 825 boyle b. j. , shanks t. , peterson b. a. , 1988 , mnras , 235 , 935 cavaliere a. , szalay a. , 1986 , apj , 311 , 589 cavaliere a. , vittorini v. , 1997 , in mller v. et al .
, 1997 , proc .
12th potsdam cosmology workshop .
, preprint chokshi a. , turner e. l. , 1992 , mnras , 259 , 421 di matteo t. , fabian a. c. , 1997 , mnras , 286 , 393 efstathiou g. p. , rees m. j. , 1988 , mnras , 230 , 5p efstathiou g. p. , bond j. r. , white s. d. m. , 1992 , mnras , 258 , 1p faber s.m .
et al . , 1997 , aj , 114 , 1771 fabian a.c . , barcons x. , almaini o. , iwasawa k. , mnras , preprint ford h.c . , tsvetanov z.i . , ferarese l. , jaffe w. , 1997 , iau symposium 184 kyoto 1997 , astro - ph/9711299 fugukita m. , hogan c.j . , peebles p.j.e .
1997 , submitted to apj , astro - ph/9712020 genzel r. , eckart a. , ott t. , eisenhauer f. , 1997 , mnras , 201 , 219 giavalisco m. , steidel c. s. macchetto f. d. , 1996 , apj , 470 , 189 haehnelt m.g .
, rees m.j .
, 1993 , mnras , 263 , 168 ( hr93 ) haehnelt m.g .
, steinmetz m. , 1997 , mnras , in press haiman z. , loeb a. , 1997 , submitted to apj , astro - ph/9710208 ( hl97 ) hasinger g. , 1998 , astron .
nachr , in press , preprint hasinger g. , burg r. , giacconi r. , schmidt j. , trmper j. , zamaroni g. , 1997 , aa in press , astro - ph/9709142 ho l.c . ,
fillipenko , a.v . ,
sargent w.l.w . ,
apj , 1997 , 487 , 568 jenkins a. et al . , 1997 , submitted to apj , astro - ph/9709010 .
jing j.p . ,
suto y. , 1997 , submitted to apj , astro - ph/9710090 kaiser n. , 1984 , apj 284 , l9 lemson g. , kauffmann g. , 1997 , submitted to mnras , astro - ph/9710125 la franca f. , andreani p. , cristiani s. , 1997,apj , in press , astro - ph/9711048 magorrian j. , et al . , 1997 , submitted to aj , astro - ph/9708072 mcmahon r. g. , irwin m. , hazard c. , 1997 , preprint mchardy , et al . , 1997 , mnras , in press , astro - ph/9703163 miyoshi m. , moran m. , hernstein j. , greenhill l. , nakai n. , diamond p. , inoue n. , 1995 , nature , 373 , 127 miralda - escud j. , 1997 , submitted to apj , astro - ph/9708253 mo h.j . ,
mao s. , white s.d.m . , 1997 , submitted to mnras , astro - ph/9707093 narayan r. , yi i. , 1995 , apj , 452 , 710 natarajan p. , sigurdsson s. , silk j. , 1997 , submitted to apj , astro - ph/9710154 navarro j. , steinmetz m. , 1997 , apj , 478 , 13 peacock j.a . , 1997 , knaw colloquium , the most distant radio galaxies , astro - ph/9712068 phinney e. s. , 1997 , talk presented at the iau symposium 186 kyoto 1997 pettini m. , steidel c.s . , adelberger k.l . , kellogg m. , dickinson m. , giavalisco m. , 1997 , in shull j.m . , woodward c.e .
, thronson h. , eds , origins , asp conference series , astro - ph/9708117 rees m. j. , 1984 , araa , 22 , 471 reimers d. , khler s. , wisotzki l. , groote d. , rodriguez - pascal p. , wamsteker w. , 1997 , aa , 327 , 890 schmidt m. , schneider d. p. , gunn j. e. , 1994 , aj , 110 , 68 schmidt m. et al .
, 1997 , aa , in press , astro - ph/9709144 shaver p. , wall j. v. , kellermann , k. i. , jackson c. a. , hawkins m. r. s. , 1996 , nature , 384 , 439 silk j. , rees m.j . , 1998 ,
a&a , in press shlosman i. , begelman m.c . , julian f. , 1990 , nat .
, 345 , 679 small t. a. , blandford r. d. , 1992 , mnras , 259 , 725 soltan a. , 1982 , mnras , 200 , 115 .
songaila a. , cowie l.l . , 1996 , aj , 112 , 335 steidel c. s. , hamilton d. , 1992 , aj , 104 , 941 steidel c. s. , pettini m. , hamilton d. , 1995 , aj , 110 , 2519 steidel c. s. , giavalisco m. , pettini m. , dickinson m. , adelberger k. l , 1996 , apj , 462 , l17 steidel c. s. , adelberger k. , dickinson m. , giavalisco m. , pettini m. , 1997 , apj , in press , astro - ph/9708125 van der marel r. p. , 1997
, in sanders d.b .
, barnes j. , eds , iau symposium 186 kyoto 1997 .
kluwer , astro - ph/9712076 wandel a. , 1991 , aa , 241 , 5 warren s.j . ,
hewitt p.c .
, osmer p.s .
, 1994 , apj , 421 , 412 watson w.d .
, wallin b.k . , 1994 ,
apj , 432 , l35 white s.d.m .
, efstathiou g. , frenk c. , 1993 , mnras , 262 , 1023 yi i. , 1996 , apj , 473 , 645 yi i. , boughn s.p .
, 1997 , submitted to apj , astro - ph/9710147 | we demonstrate that the luminosity function of the recently detected population of actively star - forming galaxies at redshift three and the b - band qso luminosity function at the same redshift can both be matched with the mass function of dark matter haloes predicted by standard variants of hierarchical cosmogonies for lifetimes of optically bright qsos anywhere in the range @xmath0 to @xmath1 .
there is a strong correlation between the lifetime and the required degree of non - linearity in the relation between black hole and halo mass .
we suggest that the mass of supermassive black holes may be limited by the back - reaction of the emitted energy on the accretion flow in a self - gravitating disc .
this would imply a relation of black hole to halo mass of the form @xmath2 and a typical duration of the optically bright qso phase of a few times @xmath3 .
the high integrated mass density of black holes inferred from recent black hole mass estimates in nearby galaxies may indicate that the overall efficiency of supermassive black holes for producing blue light is smaller than previously assumed .
we discuss three possible accretion modes with low optical emission efficiency : ( i ) accretion at far above the eddington rate , ( ii ) accretion obscured by dust , and ( iii ) accretion below the critical rate leading to an advection dominated accretion flow lasting for a hubble time .
we further argue that accretion with low optical efficiency might be closely related to the origin of the hard x - ray background and that the ionizing background might be progressively dominated by stars rather than qsos at higher redshift .
# 1#1 # 1#1 [ firstpage ] galaxies : formation , nuclei quasars : general black hole physics |
x - ray spectra of optically - thin hot plasmas in supernova remnants ( snrs ) provide key information on the nucleosynthesis during the stellar evolution and the supernova ( sn ) explosion . since iron ( fe ) is the final product of the major nuclear reaction network in a massive star and a sn
, it is particularly important element .
so far , fe abundances of some snrs have been estimated using the fe l - shell lines , which are dominant in the low energy band around 1kev .
this energy band is complex with other strong emission lines such as oxygen ( o ) and neon ( ne ) .
moreover the fe l - lines consist of many emission lines in this narrow energy band , and hence conventional x - ray detectors such as x - ray ccd can not resolve these many lines .
the fe k - shell lines at 6.46.7kev are simpler and their emission model is more reliable than those of the l - shell lines . however , due to the limited line fluxes and the sensitivity in the energy band above 6kev , the fe k - shell lines have been reported from only a limited number of x - ray bright snrs .
core - collapse ( cc ) sne eject less fe than those of type ia sne ( @xcite ; @xcite , and references therein ) .
most of the fe in the center of the progenitor star is collapsed into a central compact object ( neutron star or black hole ) .
thus observationally , fe k - shell lines would become weaker than those in type ia . the flux of fe may come from the combined effect of the elemental abundances and the thermal state of the gas .
thus the differences between type ia and cc snrs may be related to the initial condition of the circumstellar medium ( csm ) and an explosion mechanism .
fe k - shell lines have been reported from other candidates of cc snrs ( e.g. , w49b : @xcite ; ic443 : @xcite ; g349.7 + 0.2 : @xcite and g350.1 - 0.3 : @xcite ; 3c397 : @xcite )
. however , some of them are controversial whether they are cc snrs or not .
the most reliable criteria of cc sn are the presence of a neutron star ( pulsar ) , a pulsar wind nebula ( pwn ) , and oxygen - rich ( o - rich ) knots , because o is largely enhanced in the ejecta of a cc sn . at present
, three o - rich remnants , cas a , puppisa and [email protected] , have been reported in our galaxy ( see @xcite , and references therein ) .
cas a is the most luminous , and hence the fe k - shell line was firstly detected at 6.6kev ( @xcite ) .
the spatial distribution is not a simple stratified structure ( @xcite ) .
thus the sn explosion would be highly asymmetric .
puppisa has also the interesting distribution of the ejecta , in that the ejecta is found only in the east , mostly north - east portion ( @xcite ; @xcite ; @xcite ) , while a neutron star is propelled in the opposite direction ( e.g. , @xcite and references therein ) .
such a recoil between sn ejecta and a neutron star is expected in a recent sn explosion model ( @xcite ) , and would be examined by asymmetric fe distribution .
however , no fe k - shell line has been observed from puppisa . for the study of the fe k - shell line in cc snrs , we observed [email protected] .
the discovery of a pulsar and a pwn in [email protected] is further confirmation of a cc snr ( @xcite , @xcite ; @xcite ; @xcite ) .
the pulsar is located about 0.9southeast from the geometrical center of the snr .
the distance of [email protected] is estimated to be 6kpc ( @xcite ) , and the age is likely 2990@xmath260years ( @xcite ) .
the morphology of [email protected] consists of many small knots and the central belt - like filaments running from the east to the west ( @xcite ) .
the central filaments have a normal solar - type composition , suggesting that these are the shocked csm .
@xcite reproduced the intensity profile of the outer csm region by a slow wind from a red supergiant ( rsg ) star with the total mass of the wind of 1540_m_@xmath3 .
the implied progenitor mass ( @xmath4 20_m_@xmath3 ) was in plausible agreement with previous estimates ( @xcite ; @xcite ; @xcite ) .
the knots have an enhanced metallicity ; si is enhanced in north - northeast , o is enhanced primarily in southeast , ne is in northwest and southeast , and mg is in northwest ( @xcite ) .
these knots are probably ejecta origin . the asymmetric distribution of the ejecta elements is interpreted to be non - uniform thermodynamic conditions of the x - ray - emitting ejecta ( @xcite ) .
in spite of these extensive studies , no significant fe k - shell line has been detected .
@xcite proposed that the ejecta are strongly stratified by composition and the reverse shock has not propagated to the fe rich - zone yet .
however , the x - ray spectra reported so far are faint in the hard band ( except for that of pwn ) , and hence observed lines have been limited up to sulfur ( s ) k - shell line . in this paper , we report the suzaku discovery of an fe k - shell line in a high temperature plasma ( @xmath5 = 23kev ) extending to @xmath6 = 10kev .
k - shell lines of argon ( ar ) and calcium ( ca ) are also reported .
based on the wide band spectral analysis , we discuss the nature of [email protected] .
we adopt the solar abundances of @xcite . unless otherwise specified , all errors represent 1@xmath7 confidence levels .
the suzaku satellite ( @xcite ) observed [email protected] on 2011 july 22 - 23 ( obsid : 506062010 , pi : k. koyama ) with the x - ray imaging spectrometer ( xis , @xcite ) .
the xis consists of four x - ray ccd cameras placed on the focal plane of the x - ray telescope ( xrt ) .
all four xrts are co - aligned to image the same region of the sky .
the field of view ( fov ) of the xis is @xmath8 .
details of suzaku , the xis and the xrt are given in @xcite , @xcite and @xcite , respectively .
three of the xis ( xis0 , xis2 , and xis3 ) have front - illuminated ( fi ) ccds , sensitive in the 0.414kev energy band , and the other ( xis1 ) has a back - illuminated ( bi ) ccd , with high sensitivity down to 0.2kev .
xis2 has been out of function from 2006 november 9 and a small fraction of the xis0 area has not been available from 2009 june 23 , both due to the damage by micro - meteorites .
data reduction and analysis were performed by the heasoft version 6.9 .
the xis data were processed with the suzaku pipe - line software version 2.7 .
we combined the 3@xmath83 and 5@xmath85 event files .
the response functions were generated by using the caldb 2012 - 10 - 09 .
after removing hot and flickering pixels , we compiled the data using the asca - grade 0 , 2 , 3 , 4 , and 6 data .
we excluded the data obtained at the south atlantic anomaly , during the earth occultation , at the elevation angle from the earth rim below 5@xmath9 ( night earth ) and 20@xmath9 ( day earth ) .
the exposure time after these screenings was 44ks .
the spectral resolution has been degraded due to the radiation of cosmic particles 5 years after the launch , and restored by the spaced - row charge injection ( sci ) technique ; the charge traps are filled by the artificially injected electrons through ccd readouts .
details of the sci technique are given in @xcite and @xcite .
figure [ xis_image ] ( a)-(c ) shows x - ray images in the 0.30.8 , 0.86 and 68kev energy bands . in the high energy band above 6kev ( figure [ xis_image ] ( c ) ) , we see a compact x - ray source at ( @xmath10 ) = ( , ) .
this source corresponds to the pulsar / pwn ( @xcite , @xcite ; @xcite ; @xcite ) .
we make two source spectra , one is from the solid circle ( radius of ) as shown figure [ xis_image ] ( c ) ( here pwn region ) .
the other is from the solid circle ( radius of ) excluding the dashed circle ( radius of ) as shown in figure [ xis_image ] ( b ) ( here snr region ) . in figure [ xis_image ] ( a )
, we show the background ( bg ) region : the whole fov of the xis ( solid square ) , excluding the dashed circle ( radius of ) . this larger radius than that of the snr region is employed to avoid the contamination from the snr .
we also exclude the region of calibration sources , which are shown by the dashed circles in the xis corner .
for all the regions of pwn , snr and bg , we separately make the non x - ray background ( nxb ) spectra using _ xisnxbgen _ in the heasoft package ( @xcite ) .
we make an x - ray background ( xb ) spectrum from the bg region by subtracting the relevant nxb .
the spectra from the snr and pwn regions are also made by subtraction of relevant nxbs for these regions . from these spectra
, we subtract the xb spectrum assuming the uniform distribution within the fov of the xis after the correction of the vignetting effect .
the resulting spectra of the snr and the pwn regions in figure [ snr - pwn_spec ] show many emission lines . from the center energies of these lines , we identify them to be o ly@xmath11 , ne he@xmath11 , ne ly@xmath11 , mg he@xmath11 , mg ly@xmath11 , si he@xmath11 and s he@xmath11 .
thus the spectra should be composed of an optically thin hot plasma with the temperature @xmath121kev ( @xcite ; @xcite ; @xcite ) . in the spectrum of the snr region ,
we find line - like features at 3.1 , 3.9 and 6.6 kev , which are likely k - shell transition lines of ar , ca and fe . the fe line at 6.6 kev is particularly clear .
therefore , in addition to the 1kev plasma ( low-@xmath5 plasma ) , a higher temperature plasma ( high-@xmath5 plasma ) to emit k - shell lines of ar , ca and fe should be prevailing in the snr region .
we call these two plasmas the snr components
. the x - rays of the snr - components distribute not only in the snr region but also in the pwn region . on the other hand , according to the observation with chandra , the pwn is compact of
( @xcite ) , and the spectrum is fitted with a power - law model of @xmath13 1.7 ( @xcite ) , indicating its non - thermal nature .
we call this power - law emission the pwn - component .
the spectrum extracted from the pwn region contains both the snr - components and the pwn - component .
the spectrum from the snr region , on the other hand , is contaminated by x - rays of the pwn component due to the large point spread function ( beam size ) of the suzaku xrt .
we hence simultaneously fit the spectra in the snr and the pwn regions with the combined model , snr plus pwn - components .
the ancillary response files ( arfs ) employed in the fit are generated with _
xissimarfgen _ ( @xcite ) .
the arf for the pwn - component is generated from the chandra image in the 47kev band , while that for the snr - component is made using the thermal emission of the chandra image ( 0.62.0kev ) , where the emissions of the pwn - component are excluded .
the energy ranges of the pwn and the snr regions are 19kev and 0.69kev , respectively . the former energy band is selected because the contamination of the snr - component in the pwn region becomes large below 1kev .
considering the background level , we also ignore the energy band upper than 9kev for the fi , while 8kev for the bi .
@xcite and @xcite reported that the spectra are significantly different from position to position .
therefore , the integrated spectrum from the entire snr can not be described by any single component model fit .
we thus search for a many - components model , starting from one - component model then adding another component one by one , monitoring how much @xmath14 is reduced .
we use vpshock ( @xcite ) to represent multi-@xmath15 non equilibrium ionization ( nei ) plasma , where @xmath16 and @xmath17 are the plasma density and elapsed time after the shock heating . in order to fine - tune the calibration errors , between xis0 , 1 and 3 , the gains and normalizations are set to be independent parameters for each xis .
a 1-vpshock model fails with extreme large @xmath18d.o.f .
of 15006/2071 = 7.25 . a 2-vpshock model is largely improved the fitting with @xmath18d.o.f . of 5400/2060 = 2.62 , but still unacceptable .
we thus add the third vpshock component ( vpshock 1 , 2 and 3 ) , then @xmath18d.o.f .
is improved to 5076/2049 = 2.48 .
although the decrease of @xmath18d.o.f .
is only @xmath19 = 0.14 , the decrease of @xmath14 is 324 , which is statistically highly significant .
in fact , we check the significance using an f - test tool in the xspec package , then this process is significant with better than 0.01% level . although errors are large , the best - fit abundances in vpshock 1 and 2 are the same with each other .
we hence link the abundances in these two vpshock components .
also abundances of ni and ca are linked to fe and ar , respectively .
since @xmath15 of vpshock 3 is 10@xmath20 @xmath21s@xmath22 , we replace this model by an apec model ( collisional ionization equilibrium plasma model ; cie ) . the @xmath18d.o.f . of this fit
is 5184/2062 = 2.51 , leaving large residuals in the low energy band .
we thus added another apec component linking the abundances to the apec component in the 3-component model .
this another apec improve the @xmath18d.o.f . to 4838/2060 = 2.35
, the f - test statistical significant is even better than the previous process .
though this @xmath18d.o.f is still large from a statistical point of view , its value would be due to non - negligible systematic errors .
in fact , we find line - like residuals at about 0.82 , 1.2 , 1.3 and 1.8kev for both fi and bi , and 1.7kev for bi .
the 1.7 and 1.8kev line structures are due to the well - known problem of the response function near the neutral si k - edge energy at 1.84kev ( @xcite ) .
the other line structures would be due to the incompleteness of the vpshock model code .
the lines at 0.82 , 1.2 and 1.3kev correspond to fe - l complex ( @xcite ; @xcite ) .
we thus added extra 5 gaussians to compensate these line - like residuals .
the normalization factors of these gaussians are linked between fi and bi , but that of the 1.7 and 1.8kev lines are treated as an independent parameters between fi and bi ( @xcite ) .
the calibration errors of the contamination on the optical blocking filter has some problems in the low energy band ( http://www.astro.isas.jaxa.jp/suzaku/doc/suzaku_td/ ) . for a fine - tuning of the cross errors between the fi and bi ccds , we allow the @xmath23 value in the bi ccd to be independent from the fi ccds .
the results of the combined fits by this model are shown in figure [ snr - pwn_spec ] .
the best - fit parameters are given in table [ snr - pwn_para ] .
we finally improve the @xmath18d.o.f . to 2872/2059 = 1.39 .
the f - test statistical significant is better than 0.01% .
thus , we regard that this model ( 2-vpshock @xmath0 2-apec @xmath0 pl
@xmath0 5-gaussians ) is a reasonable approximation for the snr and pwn spectra , and apply in the following analysis and discussion . in order to examine spatial asymmetry of the elements in the snr
, we make spatially - resolved spectra .
since the spatial resolution of suzaku is limited compared to the size of [email protected] ( @xmath24@xmath8 ) , we crudely divide the snr into 3 : the center , north and south regions as shown in figure [ n - c - s_region ] by the solid lines .
the spectra are given in figure [ each_region_spec ] .
we find no significant differences , except a hint of fe k@xmath11 flux variations among the 3 regions . for quantitative estimate
, we fit with the same model and the method given in subsection 3.1 , but @xmath23s are fixed to the best - fit values in table [ snr - pwn_para ] .
we obtain nice fit with @xmath18d.o.f . of 1283/973 = 1.32 , 1391/1055 = 1.32 and 1380/1022 = 1.35 , for the north , center and south regions , respectively .
the best - fit spectra are given in figure [ each_region_spec ] by the solid histograms .
the best - fit parameters , including abundances , show no spatial - difference within their large errors ( typical errors for the abundances are @xmath2450% ) . the only spatial - difference is found in the fe abundances of the nei plasma ( 2-vpshock ) , which are [email protected] , [email protected] and [email protected] , for the north , center and south regions , respectively . the fe abundance in the nei plasma may be affected by the flux of the power - law component ( pwn component ) , because the continuum emission of the ejecta is equal or even less than the power - law emission except the north region ( see figure [ each_region_spec ] ) .
we therefore re - fit the spatially - resolved spectra , changing the normalizations of the pwn component by @xmath25% of the value in table [ snr - pwn_para ] , and fixing the photon index of @xmath25 = 1.91 .
the fit gives no essential changes of the fe abundances from those of the original value .
in the plasma evolution in snrs , the x - ray emissions have two different components : the csm ( plus ism ) heated by the blast wave , and the ejecta from the progenitor star heated by the reverse shock . in the spectral
fitting of [email protected] , we find two type plasmas in cie and nei conditions ; the 2-apec ( @xmath260.2 and 0.7kev ) and the 2-vpshock ( @xmath261 and 2.5kev ) plasmas .
we call these two type plasmas , the low-@xmath5 plasma and the high-@xmath5 plasma , respectively
. since the low-@xmath5 has nearly solar abundances for all elements and the high-@xmath5 has super - solar abundances ( see table [ snr - pwn_para ] ) , these would be the csm plus ism and the ejecta origin of a cc sn , respectively .
chandra spectra from many selected regions of bright small spots are described by 1-vpshock model with super - solar abundances ( @xcite ) , while those from the faint outer - most shell are 1 or 2-vpshock model with sub - solar to solar abundances ( @xcite ; @xcite ) .
these results of `` no - cie '' plasma are in contrast to the existence of cie ( @xmath24 solar ) components in the suzaku spectra .
since the chandra results are from selected spots or filaments and those of suzaku are from the entire snr region , we suspect that our cie plasmas are prevailing over the entire snr except the outer - most shell , while the bright spots are dominated by the vpshock plasma .
we discover fe k - shell line at 6.6 kev in the eject plasma for the first time .
the energy indicates that ionization state of fe is around b - like .
this medium ionization state is similar to another young cc snr , cas a , but is in contrast to nearly ne - like states in young well known type ia snrs , tycho , kepler , and sn1006 .
figure [ ejecta_abundance_pattern ] is the abundance plot of the ejecta for o , ne , mg , si , s , ar and fe relative to si ( from table [ snr - pwn_para ] ) together with those of the cc sn model in various progenitor masses ( @xcite ) .
we see that the observed abundance pattern is globally in agreement with the model of 3035_m_@xmath3 .
these mass range of the progenitor star confirm the previous report of 3040_m_@xmath3 ( @xcite ) , which was estimated based on the limited information of non - detection of the explosively synthesized heavy elements such as ar , ca and fe .
one may argue that cc sn of a massive progenitor 3035_m_@xmath3 may leave a black hole rather than a neutron star .
however , other observations show that a neutron star can be still formed from even these massive progenitor stars ( @xcite ) .
although @xcite and @xcite reported significant spatial variations in the sub - arcmin scale , we find no significant and systematic variations in the spatial scale over arcmin .
in fact , the best - fit abundances of most of the heavy elements show no variations within their typical errors of 50% .
nevertheless , we find marginal evidence of spatial variation of fe in the ejecta ; the north region is enhanced compared to that of the center region .
since the position of the neutron star ( pwn ) is off - set to southeast from the geometrical snr center ( @xcite ) , it would be conceivable that fe from the core region would be ejected to the opposite northwest direction .
our observational result of the fe variation is marginal to support this off - set effect due to large errors . to establish this kick - off scenario , we need higher quality observations . the best - fit spectral parameters of the pwn , the photon index and unabsorbed flux ( 48kev ) are [email protected] and ( [email protected])@xmath810@xmath27 photons s@xmath22 @xmath28 , respectively . the photon index is steeper than that of the pulsar ( 1.61.7 , @xcite , @xcite ) .
probably the index increases as the distance from the central pulsar increases ( e.g. @xcite ) .
the pwn flux is 52% of the total flux ( 48kev ) from the whole snr ( ( [email protected])@xmath810@xmath27 photons s@xmath22 @xmath28 ) .
this ratio is slightly smaller than 66% , determined with the high spatial resolution observation of chandra ( hughes et al .
this difference , however , would be within uncertainty range due possibly to the nxb and cxb subtraction , and/or other systematic cross errors including different data reduction processes between suzaku and chandra .
thus our simultaneous fitting analysis properly estimates the flux and spectra of both the snr and the pwn , although the spatial resolution of suzaku is limited to completely separate the emissions from these two sources .
we have analyzed suzaku / xis data obtained from [email protected] .
the results are summarized as follows : 1 .
we confirm that the thermal x - ray emission from [email protected] consists of two type plasmas in cie and nei conditions .
the nei plasma includes k - shell line from b - like fe , with super solar abundances for o , ne , mg , si , s , ar , and fe .
therefore this plasma is likely the ejecta origin of the cc snr .
3 . using the abundance pattern of the ejecta
, we confirm the progenitor mass to be 3035_m_@xmath1 .
the cie plasma has nearly solar abundances for all the relevant elements , and hence is likely the csm and ism origin .
we thank all members of the suzaku operation and calibration teams . this work is supported by japan society for the promotion of science ( jsps ) kakenhi grant number 24540229 ( katsuji koyama ) , 23000004 ( hiroshi tsunemi ) , 23340071 ( kiyoshi hayashida ) , 24684010 ( hiroshi nakajima ) , 12j01194 ( hiroaki takahashi ) , 12j01190 ( shutaro ueda ) , 24740167 ( koji mori ) , 25800119 ( satoru katsuda ) , 11j00535 ( hiroyuki uchida ) .
is also supported by the special postdoctoral researchers program in riken .
@l@l@c@ component & parameter & value + absorption & @xmath23 ( 10@xmath29 @xmath28 ) & [email protected] ( fi ) + & & [email protected] ( bi ) + power - law & photon index & [email protected] + & absorbed flux & [email protected] + & unabsorbed flux & [email protected] + apec 1 & @xmath5 ( kev ) & [email protected] + & o & [email protected] + & ne & [email protected] + & mg & [email protected] + & si & [email protected] + & s & [email protected] + & ar ( = ca ) & @xmath30 1.21 + & fe ( = ni ) & [email protected] + & vem ( 10@xmath31 @xmath32 ) & 230@xmath2160 + apec 2 & @xmath5 ( kev ) & [email protected] + & vem ( 10@xmath31 @xmath32 ) & 105@xmath215 + vpshock 1 & @xmath5 ( kev ) & [email protected] + & o & [email protected] + & ne & [email protected] + & mg & [email protected] + & si & [email protected] + & s & [email protected] + & ar ( = ca ) & [email protected] + & fe ( = ni ) & [email protected] + & @xmath15 ( 10@xmath31 @xmath21 s ) & [email protected] + & vem ( 10@xmath31 @xmath32 ) & [email protected] + vpshock 2 & @xmath5 ( kev ) & [email protected] + & @xmath15 ( 10@xmath31 @xmath21 s ) & [email protected] + & vem ( 10@xmath31 @xmath32 ) & [email protected] + @xmath18d.o.f . & & 1.39 ( 2872/2059 ) + | we report the suzaku / xis results of the galactic oxygen - rich supernova remnant ( snr ) , [email protected] , a remnant of a core - collapse supernova .
the x - ray spectrum of [email protected] consists of two type plasmas , one is in collisional ionization equilibrium ( cie ) and the other is in non - equilibrium ionization ( nei ) .
the cie plasma has nearly solar abundances , and hence would be originated from the circumstellar and interstellar mediums .
the nei plasma has super - solar abundances , and the abundance pattern indicates that the plasma originates from the supernova ejecta with a main sequence of 3035_m_@xmath1 .
iron k - shell line at energy of 6.6kev is detected for the first time in the nei plasma . |
the nonlinear delay differential equation @xmath0 where @xmath1 are positive constants , proposed by mackey and glass @xcite , has been used as an appropriate model for the dynamics of hematopoiesis ( blood cells production ) @xcite . in medical terms , @xmath2 denotes the density of mature cells in blood circulation at time @xmath3 and @xmath4 is the time delay between the production of immature cells in the bone marrow and their maturation for release in circulating bloodstream .
as we may know , the periodic or almost periodic phenomena are popular in various natural problems of real world applications @xcite . in comparing with periodicity ,
almost periodicity is more frequent in nature and much more complicated in studying for such model @xcite . on the other hand , many dynamical systems describe the real phenomena depend on the history as well as undergo abrupt changes in their states .
this kind of models are best described by impulsive delay differential equations @xcite
. a great deal of effort from researchers has been devoted to study the existence and asymptotic behavior of almost periodic solutions of and its generalizations due to their extensively realistic significance .
we refer the reader to @xcite and the references therein .
particularly , in @xcite , wang and zhang investigated the existence , nonexistence and uniqueness of positive almost periodic solution of the following model @xmath5 by using a new fixed point theorem in cone . very recently , using a fixed point theorem for contraction mapping combining with the lyapunov functional method , zhang et al .
@xcite obtained sufficient conditions for the existence and exponential stability of a positive almost periodic solution to a generalized model of @xmath6 by employing a novel argument , a delay - independent criteria was established in @xcite ensuring the existence , uniqueness , and global exponential stability of positive almost periodic solutions of a non - autonomous delayed model of hematopoiesis with almost periodic coefficients and delays . in @xcite , alzabut et al . considered the following model of hematopoiesis with impulses @xmath7 where @xmath8 represents the instant at which the density suffers an increment of @xmath9 unit and @xmath10
the density of mature cells in blood circulation decreases at prescribed instant @xmath11 by some medication and it is proportional to the density at that time @xmath12 . by employing the contraction mapping principle and applying gronwall - bellman s inequality , sufficient conditions which guarantee the existence and exponential stability of a positive almost periodic solution of system were given in @xcite as follows .
[ thm1 ] assume that * the function @xmath13 is almost periodic in the sense of bohr and there exists a positive constant @xmath14 such that @xmath15 .
* the sequence @xmath16 is almost periodic and @xmath17 . *
the sequences @xmath18 are uniformly almost periodic and there exists a positive constant @xmath19 such that @xmath20 , where @xmath21 , @xmath22 and @xmath23 . *
the function @xmath24 is almost periodic in the sense of bohr , @xmath25 , and there exists a positive constant @xmath26 such that @xmath27 . * the sequence @xmath28 is almost periodic and there exists a constant @xmath29 such that @xmath30 .
if @xmath31 , then equation has a unique positive almost periodic solution .
unfortunately , the above theorem is incorrect .
for this , let us consider the following example .
[ exam1 ] consider the following equation @xmath32 note that is a special case of .
moreover , we can easily see that equation satisfies conditions ( c1)-(c5 ) , where @xmath33 and @xmath34 .
suppose that system has a positive almost periodic solution @xmath35 .
it is obvious that @xmath36 for any positive integer @xmath37 , we have @xmath38 which yields a contradiction .
this shows that has no positive almost periodic solution .
thus , theorem [ thm1 ] is incorrect , and theorem 3.2 in @xcite is also incorrect . motivated by the aforementioned discussions , in this paper we consider a generalized model of hematopoiesis with delays , harvesting terms @xcite and impulses of the form @xmath39,\quad t\neq t_k , \end{aligned}\\ & \delta x(t_k)=x\left(t_k^+\right)-x\left(t^-_k\right ) = \gamma_kx\left(t_k^-\right)+\delta_k , \quad k\in \mathbb{z } , \end{split}\ ] ] where @xmath40 is given positive integer , @xmath41 ,
are nonnegative functions ; @xmath42 are nonnegative functions represent harvesting terms ; @xmath43 , are time delays ; @xmath44 are positive numbers and @xmath45 is a constant ; @xmath46 are constants ; @xmath47 are nonnegative integrable functions on @xmath48 $ ] with @xmath49 ; @xmath50 , is an increasing sequence involving the fixed impulsive points with @xmath51 .
the main goal of the present paper is to establish conditions for the existence of a unique positive almost periodic solution of model .
it is also proved that , under the proposed conditions , the unique positive almost periodic solution of is globally exponentially attractive .
the rest of this paper is organized as follows .
section 2 introduces some notations , basic definitions and technical lemmas .
main results on the existence and exponential attractivity of a unique positive almost periodic solution of are presented in section 3 .
an illustrative example is given in section 4 .
the paper ends with the conclusion and cited references .
let @xmath52 be a fixed sequence of real numbers satisfying @xmath53 for all @xmath54 , @xmath51 .
let @xmath55 be an interval of @xmath56 , denoted by @xmath57 the space of all piecewise left continuous functions @xmath58 with points of discontinuity of the first kind at @xmath59 , @xmath54 .
the following notations will be used in this paper . for bounded functions @xmath60 , @xmath61 and a bounded sequence @xmath62
, we set @xmath63 the following definitions are borrowed from @xcite .
[ def1 ] the set of sequences @xmath18 , where @xmath64 , is said to be uniformly almost periodic if for any positive number @xmath65 , there exists a relatively dense set of @xmath65-almost periods common for all sequences .
[ def2 ] a function @xmath66 is said to be almost periodic if the following conditions hold * the set of sequences @xmath18 is uniformly almost periodic . * for any @xmath67 , there exists @xmath68 such that , if @xmath69 belong to the same interval of continuity of @xmath70 , @xmath71 , then @xmath72 . * for any @xmath67 , there exists a relatively dense set @xmath73 of @xmath65-almost periods such that , if @xmath74 then @xmath75 for all @xmath76 , @xmath54 satisfying @xmath77 . for equation ,
we introduce the following assumptions * the function @xmath78 is almost periodic in the sense of bohr and @xmath79 . *
the functions @xmath80 , are nonnegative and almost periodic in the sense of bohr . * the function @xmath81 ,
are bounded nonnegative and almost periodic in the sense of bohr in @xmath76 uniformly in @xmath82 and there exist positive constants @xmath83 such that @xmath84 * the functions @xmath85 , are almost periodic in the sense of bohr , @xmath86 are bounded , @xmath87 , @xmath88 . *
the sequence @xmath28 is almost periodic . *
the sequence @xmath16 is almost periodic satisfying @xmath89 where @xmath90 . *
the set of sequences @xmath18 is uniformly almost periodic , @xmath91 .
[ rm1 ] it should be noted that model includes as a special case . for that model , assumptions ( a3 ) , ( a4 )
obviously be removed .
furthermore , we make assumption ( a6 ) in order to correct condition ( c2 ) in @xcite . the following lemmas will be used in the proof of our main results .
[ lem1 ] let assumption ( a7 ) holds .
assume that functions @xmath92 , are almost periodic in the sense of bohr , a function @xmath70 and sequences @xmath28 , @xmath16 are almost periodic .
then for any @xmath67 , there exist @xmath93 , relatively dense sets @xmath94 , @xmath95 such that * @xmath96 ; * @xmath97 ; * @xmath98 [ lem2 ] let assumption ( a7 ) holds .
assume that functions @xmath99 , are almost periodic in @xmath76 uniformly in @xmath100 in the sense of bohr , a function @xmath70 and sequences @xmath28 , @xmath16 are almost periodic . then for any compact set @xmath101 and positive number @xmath65
, there exist @xmath93 , relatively dense sets @xmath94 , @xmath95 such that * @xmath102 ; * @xmath103 ; * @xmath98 the proof of this lemma is similar to the proof of lemma 2.1 in @xcite so let us omit it here .
[ lem3 ] for given @xmath104 , real number @xmath105 and integers @xmath106 such that @xmath107 , if @xmath108 for all @xmath109 and @xmath110 then @xmath111 .
the proof is straight forward , so let us omit it here .
[ lem4 ] let assumptions ( a6 ) and ( a7 ) hold . if @xmath112 satisfies @xmath113 for all @xmath109 , then @xmath114 using the facts that @xmath115 and @xmath116 , from ( a6 ) , we have @xmath117 the proof is completed .
[ lem5 ] let assumption ( a7 ) holds . for any @xmath118 , @xmath119
, we have @xmath120 the proof follows by some direct estimates and , thus , is omitted here .
now , let @xmath121 , then @xmath122 . from biomedical significance , we only consider the initial condition @xmath123,\mathbb{r}).\ ] ]
it should be noted that problem and has a unique solution @xmath124 defined on @xmath125 which is piecewise continuous with points of discontinuity of the first kind , namely @xmath126 , at which it is left continuous and the following relations are satisfied @xcite @xmath127 related to , we consider the following linear equation @xmath128 [ lem6 ] let assumptions ( a1 ) , ( a6 ) and ( a7 ) hold .
then @xmath129 where @xmath130 is the cauchy matrix of , @xmath131 and @xmath132 .
the proof is straight forward from , so let us omit it here .
similar to lemma 36 in @xcite and lemma 2.6 in @xcite we have the following lemma .
[ lem7 ] let assumptions ( a1 ) , ( a6 ) and ( a7 ) hold . then , for given @xmath133 , relatively dense sets @xmath134 , @xmath135 , satisfying * @xmath136 ; * @xmath137 for any @xmath74 , @xmath138 satisfying @xmath139 , @xmath140 , where @xmath141\right\}.\ ] ] we divide the proof into two possible cases as follows . _ case 1 : _ @xmath142 . by lemma [ lem3 ] , @xmath143 . since @xmath144 , @xmath145 , @xmath146 and @xmath147 , it follows from , that @xmath148 _ case 2 : _
similarly , we have @xmath150 by lemma [ lem4 ] , from - we obtain @xmath151 the proof is completed .
it is worth noting that , the proof of lemma [ lem7 ] is different from those in @xcite . by employing lemma [ lem4 ] , we obtain a new bound for constant @xmath152 given in .
[ lem8 ] assume that there exist constants @xmath153 , @xmath154 and a function @xmath155 satisfying * @xmath156 for @xmath157 , where @xmath158 and @xmath159 ; * @xmath160 for @xmath161 , where @xmath162 and @xmath163 denotes the upper - right dini derivative ; * @xmath164 for all @xmath54 satisfies @xmath157 .
then @xmath165 where @xmath166
let us set @xmath167 and @xmath168 .
we define an operator @xmath169 as follows @xmath170 it can be verified that @xmath171 is an almost periodic solution on @xmath172 of if and only if @xmath173 .
we define the following constants @xmath174 [ lem9 ] let assumptions ( a1)-(a7 ) hold .
if @xmath175 then @xmath176 let @xmath177 .
by lemma [ lem5 ] and lemma [ lem6 ] , from we have @xmath178 for each @xmath76 , let @xmath179 be an integer such that @xmath180 .
if @xmath181 then , by lemma [ lem6 ] , it follows from the fact @xmath182 that @xmath183 if @xmath184 then from ( a7 ) and lemma [ lem6 ] , we have @xmath185 from and we obtain @xmath186 the proof is completed .
now we are in position to introduce our main results as follows .
[ thm2 ] under the assumptions ( a1)-(a7 ) , if @xmath187 then @xmath188 is almost periodic .
let @xmath177 . for given @xmath189 ,
there exists @xmath190 such that , if @xmath69 belong to the same interval of continuity of @xmath191 then @xmath192 by lemma [ lem2 ] and lemma [ lem7 ] , there exist @xmath193 , relatively dense sets @xmath94 , @xmath135 such that , for all @xmath194 , we have @xmath195 , i\in\underline{m};\\ & |a(t+\omega)-a(t)|<\delta , \ t\in \mathbb{r};\\ & |b_i(t+\omega)-b_i(t)|<\delta,\ ; |c_i(t+\omega)-c_i(t)|<\delta , \
t\in \mathbb{r } , i\in\underline{m};\\ & |\tau_i(t+\omega)-\tau_i(t)|<\delta , \ ; |\sigma_i(t+\omega)-\sigma_i(t)|<\delta , \
t\in \mathbb{r } , i\in\underline{m};\\ & |\gamma _ { k+p}-\gamma _ k| < \delta,\ ; |\delta _ { k+p}-\delta _ k| < \delta,\ ; \end{aligned}\ ] ] let @xmath74 , @xmath196 .
one can easily see that @xmath197 we define @xmath198 . for @xmath199 , @xmath200 , let us set @xmath201 then we have @xmath202 and @xmath203 .
it can be seen from and that @xmath204 by lemma [ lem5 ] and lemma [ lem6 ] , from , and the fact that @xmath205 , we have @xmath206 and @xmath207 it should be noted that , by ( a4 ) , @xmath208 , @xmath200 , are strictly increasing functions , and thus , there exist the inverse functions @xmath209 of @xmath208 .
for each @xmath210 , denote @xmath211 then @xmath212 let @xmath213 , @xmath214 , then , by ( a4 ) , @xmath215 therefore , @xmath216 and hence , from equations - , we have @xmath217 } \left[\tau^*_i ( t_k+\omega+\epsilon)-\tau^*_i(t_k+\omega-\epsilon)\right]\\ & \leq\frac{a\epsilon}{a_l}+4a\|\phi\|\overline{\lambda}_i\epsilon \sum_{t_k<\bar t}e^{-a_l{\underline \lambda}_i(\bar t - t_k ) } \leq \bigg(\frac{1}{a_l}+\frac{4\overline{\lambda}_i\|\phi\|}{1-e^{-a_l{\underline\lambda}_i\eta}}\bigg)a\epsilon . \end{aligned}\ ] ] by the same arguments used in deriving , we obtain @xmath218 combining - , and , we readily obtain @xmath219\epsilon.\ ] ] next , let us set @xmath220 it follows from and that @xmath221 by lemmas [ lem5 ] and [ lem6 ] , from we have @xmath222 and @xmath223 where @xmath224 . from - ,
we readily obtain @xmath225\epsilon.\ ] ] now , we define @xmath226 then , from and , we have @xmath227 also using lemma [ lem5 ] and lemma [ lem6 ] , from and the fact that @xmath205 , we obtain @xmath228 let @xmath229 . similarly to and , we readily obtain @xmath230 inequalities - yield @xmath231 let us set @xmath232 then @xmath233 for each @xmath234 , there exists a unique integer @xmath235 such that @xmath236 . by lemma [ lem3 ]
we have @xmath237 .
thus @xmath238 where @xmath239 note that @xmath240 and @xmath241 , from and lemma [ lem5 ] , we have @xmath242 similarly , we obtain @xmath243 it follows from - that @xmath244\epsilon.\ ] ] we also have @xmath245 therefore @xmath246 + \frac{a\epsilon}{1-e^{-\frac{1}{2}a_l\eta}}.\ ] ] we can see clearly from , , , and that there exists a positive constant @xmath247 such that @xmath248 , for all @xmath249 , @xmath54 .
this shows that @xmath250 is almost periodic .
the proof is completed .
[ thm3 ] let assumptions ( a1)-(a7 ) hold .
if @xmath251 , defined in , is positive and @xmath252 where @xmath253 , then equation has a unique positive almost periodic solution .
we define @xmath254 .
it is worth noting that , from lemma [ lem9 ] , theorem [ thm2 ] and the assumption @xmath255 , we have @xmath256 . for any @xmath257 ,
applying lemma [ lem6 ] we obtain @xmath258 therefore @xmath259 which yields @xmath260 is a contraction mapping on @xmath261 by condition
. then @xmath260 has a unique fixed point in @xmath262 , namely @xmath263 .
it should be noted that , @xmath264 , and hence , @xmath260 also has a unique fixed point @xmath263 in @xmath265 .
this shows that has a unique positive almost periodic solution @xmath266 .
the proof is completed .
[ thm4 ] let assumptions ( a1)-(a7 ) hold . if @xmath267 and @xmath268 then has a unique positive almost periodic solution @xmath35 . moreover , every solution @xmath269 of converges exponentially to @xmath35 as @xmath270
. by theorem [ thm3 ] , has a unique positive almost periodic solution @xmath35 .
let @xmath269 be a solution of and .
we define @xmath271 then @xmath272\\ & \leq -a_lv(t)+
\sum_{i=1}^m\left(b_{im}k^*_i+c_{im}g^*_i + l_i\right)\overline{v}(t ) , \ ; t\ne t_k,\ t\geq \alpha,\\ \delta v(t_k)&=\gamma_k v(t^-_k ) , \ ; t_k\geq\alpha , k\in \mathbb{z},\end{aligned}\ ] ] where @xmath273 . by lemma [ lem8 ] ,
there exists a positive constant @xmath274 such that @xmath275 this shows that @xmath2 converges exponentially to @xmath35 as @xmath270 .
the proof is completed . the existence and exponential stability of positive almost periodic solution of
is presented in the following corollary as an application of our obtained results with @xmath276 .
[ cr1 ] under the assumptions ( a1 ) , ( a2 ) ( with @xmath277 ) and ( a5)-(a7 ) , if @xmath278 and @xmath279 where @xmath280 , then has a unique positive almost periodic solution which is exponentially stable .
in this section we give a numerical example to illustrate the effectiveness of our conditions . for illustrating purpose ,
let us consider the following equation @xmath281 where @xmath282 it should be noted that , the functions @xmath283 and @xmath284 are almost periodic in the sense of bohr , @xmath285 is almost periodic in @xmath210 uniformly in @xmath286 , @xmath287 .
therefore , assumptions ( a1)-(a5 ) and ( a7 ) are satisfied . on the other hand , @xmath288 for any @xmath289 .
thus , @xmath290 and assumption ( a6 ) is satisfied .
taking some computations we obtain @xmath291 and @xmath292 . by theorem [ thm3 ]
, equation has a unique positive almost periodic solution @xmath35 .
furthermore , it can be seen that @xmath293 , and hence , @xmath294 . by theorem [ thm4 ] , every solution @xmath295 of converges exponentially to @xmath35 as @xmath3 tends to infinity . as presented in figure 1 , state trajectories of with different initial conditions
converge to the unique positive almost periodic solution of .
this paper has dealt with the existence and exponential attractivity of a unique positive almost periodic solution for a generalized model of hematopoiesis with delays and impulses . using the contraction mapping principle and a novel type of impulsive delay inequality ,
new sufficient conditions have been derived ensuring that all solutions of the model converge exponentially to the unique positive almost periodic solution .
t. t. anh , existence and global asymptotic stability of positive periodic solutions of a lotka - volterra type competition systems with delays and feedback controls , _ electron . j. diff .
_ , vol . 2013 ( 2013 ) , no .
261 , 116 . | in this paper , a generalized model of hematopoiesis with delays and impulses is considered . by employing the contraction mapping principle and a novel type of impulsive delay inequality ,
we prove the existence of a unique positive almost periodic solution of the model . it is also proved that , under the proposed conditions in this paper , the unique positive almost periodic solution is globally exponentially attractive .
a numerical example is given to illustrate the effectiveness of the obtained results .
[ section ] [ theorem]corollary [ section ] [ section ] [ section ] [ section ] |
in the past ten years there has been a flurry of activity concerning the interaction between symplectic geometry and spectral theory .
the potential of this interaction was already pointed out by colin de verdire @xcite in a quite general setting : pseudodifferetial operators on cotangent bundles .
this paper deals with the question : can one recover a classical system ( described by symplectic geometry ) from the spectrum of its quantization ( described by spectral theory ) ?
this is a classical question in inverse spectral theory going back to pioneer works of colin de verdire and guillemin - sternberg , in the 1970s and 1980s .
many contributions followed their works , eg iantchenko sjstrand zworski @xcite .
a few global spectral results have also been obtained recently , for instance by hezari and zelditch for symmetric domains in @xmath0 @xcite or by v ngc @xcite for one degree of freedom pseudodifferential operators .
recently , the first and third author settled , jointly with l. charles , this problem for toric integrable systems @xcite .
the paper @xcite gives a full description of the semiclassical spectral theory of toric integrable systems . as a corollary of this theory , that paper solves the isospectral theorem for toric systems .
we refer to section [ sec : toric ] for more details , and to @xcite for a more complete list of references on inverse spectral results in symplectic geometry . for
an interesting semiclassical recent work see the article of guillemin , paul , and uribe @xcite on trace invariants .
the goal of the present paper is to introduce a minimalistic notion of semiclassical quantization and use it to provide a simple constructive argument proving that from the joint spectrum of a quantum system one can recover the convex hull of the classical spectrum .
this applies not only to toric integrable systems , but much more generally .
moreover , our results are general enough to apply to pseudodifferential and berezin - toeplitz quantization .
[ [ structure - of - the - paper . ] ] _ structure of the paper_. + + + + + + + + + + + + + + + + + + + + + + + + + in section [ results ] we quickly state our main results in the most important particular cases : pseudodifferential and berezin - toeplitz quantization : theorem [ theo : toeplitz ] and theorem [ theo : pseudodifferential ] . in section [ sec - prel - selfad ]
we recall some preliminaries on self - adjoint operators .
in section [ sec : quantizable ] we introduce a general quantization setting and state results that encompass theorem [ theo : toeplitz ] and theorem [ theo : pseudodifferential ] . in section [ sec :
limits ] we prove a key lemma concerning spectral limits of semiclassical operators .
section [ sec : convexity ] reviews properties of support functions and convex sets which are needed for the proofs . in section [ sec : proof ] we combine the previous sections to complete the proofs . in sections [ sec :
toeplitz ] and [ sec : pseudodifferential ] we briefly review the berezin - toeplitz quantization and @xmath1-pseudo - differential calculus and show that they are covered by the quantization procedure of section [ sec : quantizable ] .
we present a number of examples illustrating our results : section [ sec : toric ] explains how to use theorem [ theo : toeplitz ] to conclude , quickly , the isospectrality theorem for toric systems which appeared in a previous work @xcite of the first and third author and l. charles . in section [ sec - examples-1 ]
we apply our results to a physically interesting system , coupled angular momenta on @xmath2 described by d.a .
sadovski and b.i .
zhilinski in @xcite .
section [ subsec - rot ] deals with @xmath1-pseudodifferential quantization of a particle on the plane in a rotationally symmetric potential . in the last section
we extend our results to nonnecessarily commuting operators .
for simplicity , in this section we will only state our results in the two ( otherwise rather general ) cases of pseudodifferential and berezin - toeplitz quantization . in the statements below
, we fix a quantization scheme for a symplectic manifold @xmath3 and focus on a collection @xmath4 of mutually commuting self - adjoint semi - classical operators .
these operators depend on the planck constant @xmath5 , where @xmath6 is a subset of @xmath7 $ ] that accumulates at @xmath8 , and act on a hilbert space @xmath9 .
let @xmath10 are pair - wise commuting selfadjoint ( not necessarily bounded ) operators on a hilbert space @xmath11 with a common dense domain @xmath12 such that @xmath13 for all @xmath14 . the _ joint spectrum _ of @xmath15 is by definition the support of the joint spectral measure .
it is denoted by @xmath16 . for instance
, if @xmath17 s are endomorphisms of a finite dimensional vector space , then the joint spectrum of @xmath10 is the set of @xmath18 such that there exists a non - zero vector @xmath19 for which @xmath20 if @xmath21 are pairwise commuting semiclassical operators , then of course the joint spectrum of @xmath10 depends on @xmath1 . ) of a quantum toric system @xcite.,scaledwidth=80.0% ] following the physicists , we shall call _ classical spectrum _ of @xmath22 the image @xmath23 , where @xmath24 is the map of principal symbols of @xmath25 .
[ theo : toeplitz ] let @xmath3 be a prequantizable closed symplectic manifold .
let @xmath26 and let @xmath27 be a family of pairwise commuting selfadjoint berezin - toeplitz operators on @xmath3 .
let @xmath28 be the classical spectrum of @xmath29 .
then @xmath30 where the limit convergence is in the hausdorff metric . in the @xmath1-pseudodifferential
setting ( see for instance @xcite ) , the analysis is more complicated due to the possible unboundedness of the operators .
here we work with the standard hrmander class of symbols @xmath31 on @xmath32 or @xmath33 with closed @xmath34 imposing the following restriction on the dependence of @xmath35 on @xmath1 : @xmath36 where all @xmath37 are uniformly ( in @xmath1 ) bounded and supported in the same compact set .
the principal symbol @xmath38 can be unbounded .
we shall say that @xmath35 _ mildly depends on @xmath1_. fortunately , in a number of meaningful examples one deals with @xmath1-independent symbols , see e.g. section [ subsec - rot ] below .
[ theo : pseudodifferential ] let @xmath34 be either @xmath39 , @xmath40 , or a closed manifold .
let @xmath26 and let @xmath41 be a family of pairwise commuting selfadjoint @xmath1-pseudodifferential operators on @xmath34 whose symbols mildly depend on @xmath1 . then the following hold .
* from the family @xmath42 one can recover the convex hull of the classical spectrum of @xmath29 . *
if , in addition , the principal symbols of @xmath17 are bounded for every @xmath43 , the joint and the classical spectra are related by equation .
the assumption on mild dependence of the symbol on @xmath1 can not be completely dropped , see remark [ remark - hbar - dependence ] below .
an immediate consequence of the theorems is the following simple statement .
[ cor : im0 ] let @xmath25 be either : * commuting self - adjoint berezin - toeplitz operators on a prequantizable closed symplectic manifold , or * @xmath1-pseudodifferential self - adjoint operators on @xmath39 or a closed manifold whose symbols mildly depend on @xmath1 .
if the classical spectrum of @xmath22 is convex , then the joint spectrum recovers it .
first let us recall some elementary results from operator theory . for any ( not necessarily bounded )
selfadjoint operator @xmath44 on a hilbert space with a dense domain and with spectrum @xmath45 , we have @xmath46 this follows for instance from ( * ? ? ?
* proposition 5.12 ) .
we give the proof here for the reader s convenience .
if @xmath47 , then by the weyl criterium , there exists a sequence @xmath48 with @xmath49 such that @xmath50 therefore , @xmath51 , which implies that @xmath52 conversely , if @xmath45 lies in @xmath53 $ ] , then by the spectral theorem @xmath54 which yields @xmath55 this proves .
let us mention also that it implies @xmath56
we shall prove theorem [ theo : toeplitz ] and theorem [ theo : pseudodifferential ] in a more general context of semiclassical quantization .
let @xmath3 be a connected manifold ( either closed or open ) .
let @xmath57 be a subalgebra of @xmath58 containing the constants and all compactly supported functions .
we fix a subset @xmath59 $ ] that accumulates at @xmath8 .
if @xmath11 is a complex hilbert space , we denote by @xmath60 the set of linear ( possibly unbounded ) selfadjoint operators on @xmath11 with a dense domain . by a slight abuse of notation , we write @xmath61 for the _ operator norm _ of an operator , and @xmath62 for the _ uniform norm _ of a function on @xmath3 .
[ bt ] a _ semiclassical quantization _ of @xmath63 consists of a family of complex hilbert spaces @xmath9 , and a family of @xmath64-linear maps @xmath65 satisfying the following properties , where @xmath66 and @xmath67 are in @xmath57 : 1 . [ item : one ] @xmath68 * ( normalization ) * ; 2 .
[ item : garding ] for all @xmath69 there exists a constant @xmath70 such that @xmath71 * ( quasi - positivity ) * ; 3 .
[ item : norm ] let @xmath72 such that @xmath73 and has compact support , then @xmath74 * ( non - degeneracy ) * ; 4 .
[ item : symbolic ] if @xmath67 has compact support , then for all @xmath66 , @xmath75 is bounded , and we have @xmath76 * ( product formula)*. a _ quantizable _ manifold is a manifold for which there exists a semiclassical quantization .
we shall often use the following consequence of these axioms : for a bounded function @xmath66 , the operator @xmath77 is bounded . indeed ,
if @xmath78 for some @xmath79 , ( q[item : one ] ) and ( q[item : garding ] ) yield @xmath80 since our operators are selfadjoint , this implies by formula above @xmath81 next , consider the algebra @xmath82 whose elements are collections
@xmath83 , @xmath84 with the following property : for each @xmath85 there exists @xmath86 so that @xmath87 where the sequence @xmath88 is uniformly bounded in @xmath1 and supported in the same compact set @xmath89 .
the function @xmath90 is called the _ principal part _ of @xmath85 .
if @xmath90 is compactly supported as well , we say that @xmath85 is _
compactly supported_. we define a map @xmath91 _ a semiclassical operator _ is an element in the image of @xmath92 .
given @xmath93 , the function @xmath94 defined by is called _ the principal symbol _ of @xmath95 . by @xmath96
this together with the product formula ( q[item : symbolic ] ) readily yields that for every @xmath97 with compact support and every @xmath85 , @xmath98 now we are ready to show that _ the principal symbol of a semiclassical operator is unique_. indeed , if @xmath99 , then for any compactly supported function @xmath100 , we get by @xmath101 and then by , @xmath102 . by ( q[item : norm ] )
we conclude that @xmath103 . since @xmath100 is arbitrary , @xmath104 .
important examples of semiclassical quantization in the sense of definition [ bt ] are provided by symplectic geometry .
they include berezin - toeplitz quantization on closed prequantizable symplectic manifolds and ( certain versions of ) @xmath1-pseudodifferential calculus on cotangent bundles . in the latter case
the algebra @xmath105 is the usual hrmander class of symbols , while its deformation @xmath106 is more special and is defined as above .
the details will be explained in sections [ sec : toeplitz ] and [ sec : pseudodifferential ] .
the definition of semiclassical quantization above is minimalistic " : we list only those axioms which will enable us to reconstruct the classical spectrum from the quantum one , which is the main objective of the paper . in particular
, we do not require a stronger form of the correspondence principle which links together commutators of semiclassical operators with poisson brackets of their symbols , and even do not assume that the manifold @xmath3 is symplectic . for the following results recall the notion of quantizable manifold we use ( definition [ bt ] ) .
[ theo : general ] let @xmath3 be a quantizable manifold .
let @xmath26 and let @xmath107 be pairwise commuting semiclassical operators on @xmath3 .
let @xmath108 be a subset of @xmath6 that accumulates at @xmath8 .
then from the family @xmath109 one can recover the convex hull of the classical spectrum of @xmath110 .
we can strengthen theorem [ theo : general ] in case the principal symbols are bounded to obtain a uniform convergence in the hausdorff distance .
recall that the _ hausdorff distance _ between two subsets @xmath44 and @xmath111 of @xmath112 is @xmath113 where for any subset @xmath34 of @xmath112 , the set @xmath114 is @xmath115 ( see eg .
@xcite ) .
[ theo : bounded ] let @xmath3 be a quantizable manifold .
let @xmath26 and let @xmath107 be pairwise commuting semiclassical operators on @xmath3 .
let @xmath108 be a subset of @xmath6 that accumulates at @xmath8 .
assume that the principal symbols of @xmath17 , @xmath116 , are bounded .
let @xmath28 be the classical spectrum of @xmath117 .
then @xmath118 where the limit convergence is in the hausdorff metric .
theorem [ theo : general ] and theorem [ theo : bounded ] together with results of sections [ subsec - bt - prelim ] and [ subsec - pdo - prelim ] below readily yield theorems [ theo : toeplitz ] and [ theo : pseudodifferential ] of the introduction .
let @xmath119 be a uniformly ( in @xmath1 ) bounded selfadjoint operator with principal symbol @xmath66 . applying theorem [ theo : bounded ] with @xmath120 , we get that @xmath121 \,=\ , { \mathcal{s } } = f(m),\ ] ] where @xmath122 and @xmath123 are the minimum and maximum of the spectrum of @xmath119 , respectively .
since by the operator norm of a bounded selfadjoint operator equals @xmath124 , we get that the axioms of semiclassical quantization listed in definition [ bt ] above imply the following automatic " refinement of non - degeneracy axiom ( q[item : norm ] ) : @xmath125 for every bounded function @xmath126 .
[ cor : im ] if the classical spectrum is convex , then the joint spectrum recovers it .
corollary [ cor : im ] is an immediate consequence of theorem [ theo : general ] . note that we do nt need to know the precise structure of the joint spectrum of the quantum system in order to recover the classical spectrum
it suffices to know the convex hull of this joint spectrum , as a subset of @xmath127 .
fix a semiclassical quantization the sense of definition [ bt ] on a manifold @xmath3 .
[ lem : max ] take any @xmath128 with principal part @xmath90 , and let @xmath129 be the corresponding semiclassical operator .
let @xmath130 denote the supremum of the spectrum of @xmath131
. then @xmath132 for clarity we divide the proof into several steps .
[ [ step-1 . ] ] _ step 1_. + + + + + + + + + by and @xmath133 therefore it suffices to prove the lemma assuming that @xmath134 for all @xmath1 . from now on
, this will be the standing assumption until the end of the proof .
further , fix @xmath135 sufficiently small .
we claim that @xmath136 for all @xmath1 sufficiently small
. indeed , if @xmath90 is unbounded from above there is nothing to prove .
if @xmath90 is bounded from above , this follows from .
[ [ step-2 . ] ] _ step 2_. + + + + + + + + + put @xmath137 let @xmath138 be a compact set with non - empty interior and such that @xmath139 let @xmath140 be a smooth function which is identically @xmath8 outside of @xmath138 , identically @xmath141 in a compact @xmath142 with non - empty interior .
then the function @xmath143 is @xmath8 outside of @xmath138 , and , inside of @xmath138 , it is greater than or equal to @xmath8
. then , by axiom ( q[item : garding ] ) , there exists a positive constant @xmath144 such that @xmath145 as above , in what follows @xmath146 denotes a positive constant which does not depend on @xmath1 and whose value may vary from step to step .
[ [ step-3 . ] ] _ step 3_. + + + + + + + + + we claim that there exists @xmath147 such that @xmath148 and @xmath149 indeed , let @xmath150 be supported on @xmath151 and non identically @xmath8 . by axiom ( q[item : norm ] ) there exists a constant @xmath152 such that @xmath153 .
therefore there exists some @xmath154 with @xmath155 and such that @xmath156 now let @xmath157 by the product formula ( axiom ( q[item : symbolic ] ) ) we know that @xmath158 and therefore @xmath159 in the second equality above we use .
since @xmath160 , it follows from equation ( [ eq:3 ] ) that @xmath161 which proves the claim ( [ eq : u ] ) .
[ [ step-4 . ] ] _ step 4_. + + + + + + + + + put @xmath162 , where @xmath163 is from step 3 . by selfadjointness and axiom ( q[item : symbolic ] ) @xmath164 where the last inequality follows from . using ( q[item : one ] ) and the fact that @xmath165 by , we conclude that @xmath166 now , if @xmath1 is small enough then @xmath167 , and hence , in view of , @xmath168 if @xmath169 and @xmath170 if @xmath171 .
since @xmath172 is arbitrary , this together with implies ( in both cases ) that @xmath173 as required
in the case of @xmath174 dimensions.,scaledwidth=60.0% ] let @xmath175 be a closed set .
let @xmath176 be the unit sphere in @xmath127 .
the map @xmath177 given by @xmath178 is called _ the support function _ of @xmath111 ( here we deviate a little bit from the standard definition where @xmath179 is defined on the whole @xmath112 ) .
the following facts are well known ( see e.g. ( * ? ? ?
* section 7.2 ) ) : [ lem : duality ] * @xmath180 . * @xmath181 .
[ prop : phi ] @xmath182 * let @xmath44 and @xmath111 be closed sets . then we have the following equivalence : + @xmath183 * let @xmath44 be a convex closed set .
let @xmath172 .
then @xmath184 * let @xmath111 be a compact set , and let @xmath185 such that for all @xmath186 .
then @xmath179 is @xmath187-lipschitz . *
( i ) * if @xmath188 , then lemma [ lem : duality](i ) gives @xmath189 .
conversely , if @xmath190 , then by definition of the maps @xmath191 , we have @xmath192 .
we conclude by lemma [ lem : duality](ii ) . *
( ii ) * if @xmath193 and @xmath194 , we have @xmath195 and hence @xmath196 .
conversely , note that @xmath197 .
therefore @xmath198 which concludes the proof . *
( iii ) * if @xmath199 , we have @xmath200 which easily implies @xmath201
[ lem : btcommuting ] let @xmath25 be pairwise commuting selfadjoint ( possibly unbounded ) operators on a hilbert space .
for any @xmath202 , let @xmath203 be the spectrum of @xmath204 .
then for any fixed @xmath205 , @xmath206 gives the spectrum of @xmath207,scaledwidth=60.0% ] let @xmath208 be the spectral measure of @xmath17 , and let @xmath209 be the joint spectral measure on @xmath112 . for a given @xmath210 define a linear functional @xmath211 by @xmath212 .
observe that @xmath213 where @xmath214 is the push - forward of @xmath215 to @xmath216 . by definition of the support of a spectral measure
, @xmath217 is a dense subset of @xmath218 .
thus @xmath219 is a dense subset in @xmath203 .
this proves the first equality in .
the second equality follows from lemma [ lem : duality](ii ) .
now we are ready to give a proof of theorem [ theo : general ] and theorem [ theo : bounded ] .
[ [ proof - of - theorem - theogeneral . ] ] _ proof of theorem [ theo : general]_. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we denote by @xmath220 the joint spectrum of @xmath110 .
let @xmath221 and let @xmath222 . by lemma [ lem : max ] , we have @xmath223 which , in view of , reads @xmath224 therefore , the map @xmath225 can be recovered from the joint spectrum @xmath220 ; by lemma [ lem : duality ] we can then recover @xmath226 , which proves theorem [ theo : general ] .
[ [ proof - of - theoremtheobounded . ] ] _ proof of theorem [ theo : bounded ] . _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + if the principal symbols of @xmath25 are bounded then the joint spectrum @xmath220 is bounded , and it follows from proposition [ prop : phi](iii ) that the family of maps @xmath227}\ ] ] is uniformly equicontinuous for @xmath228 small enough .
this uniform equicontinuity and the compactness of the sphere @xmath176 imply that the pointwise limit is in fact uniform : @xmath229 now proposition [ prop : phi](i),(ii ) gives the inclusions @xmath230 and @xmath231 in other words , if @xmath232 , the hausdorff distance between the joint spectrum of @xmath233 and @xmath234 is less than @xmath235 , which proves theorem [ theo : bounded ] .
in this section we are concerned with berezin - toeplitz operators ( simply called in the sequel toeplitz operators ) and quantization of classical systems given by such operators , see kostant @xcite , souriau @xcite , and berezin @xcite , as well as the book @xcite by boutet de monvel and guillemin for the corresponding microlocal analysis . many well known results for pseudodifferential operators are now known for toeplitz operators , see @xcite .
let @xmath236 be a closed symplectic manifold whose symplectic form represents an integral de rham cohomology class of @xmath3 times @xmath237 . in what follows such symplectic manifolds will be called _
prequantizable_. when a prequantizable symplectic manifold @xmath236 is khler with respect to a complex structure @xmath108 , the berezin - toeplitz quantization can be given by the following geometric construction .
choose a holomorphic hermitian line bundle @xmath238 over @xmath3 so that the curvature of its ( unique ) hermitian connection compatible with the holomorphic structure equals @xmath239 ( the existence of such a @xmath238 is a well known fact from complex algebraic geometry ) .
for a positive integer @xmath240 , @xmath241 is the space of holomorphic sections of @xmath242 .
since @xmath3 is compact , @xmath243 is a finite dimensional subspace of the hilbert space @xmath244 . here
the scalar product is defined by integrating the hermitian pointwise scalar product of sections against the liouville measure of @xmath3 .
denote by @xmath245 the orthogonal projector of @xmath246 onto @xmath247 . put @xmath248 and define the quantization map @xmath249 by @xmath250 where @xmath251 is the operator of multiplication by @xmath66 . here
the planck constant @xmath1 runs over the set @xmath252 .
the fact that @xmath249 is a semiclassical quantization in the sense of definition [ bt ] is proved in @xcite . in what follows
we shall use that the berezin - toeplitz quantization behaves in a functorial way with respect to direct products of closed khler manifolds : given prequantum bundles @xmath253 and @xmath254 over @xmath255 and @xmath256 respectively , form a prequantum bundle @xmath257 over @xmath258 , where @xmath259 is the natural projection . by a version of knneth formula @xcite @xmath260 . for a pair of functions @xmath261 ,
@xmath262 define a new function @xmath263 by @xmath264 one can readily check that @xmath265 if @xmath3 is a prequantizable closed symplectic ( but not necessarily khler ) manifold , there exist several constructions of semiclassical quantizations of @xmath3 satisfying definition [ bt ] , see @xcite .
now we are ready to present some specific applications of our main results in the context of the berezin - toeplitz quantization .
assume that @xmath3 is a prequantizable khler manifold endowed with a hamiltonian @xmath266-action which preserves the complex structure .
then the kostant - souriau formula yields commuting toeplitz operators whose principal symbols are the components of the @xmath266-momentum map . a proof when @xmath267 was outlined in charles - pelayo - v ngc (
* theorem 1.4 ) as a byproduct of the main result of the paper ( a normal form theorem for quantum toric systems ) ; the @xmath268 case was stated in charles @xcite without further details
. therefore such a quantum @xmath266 action , @xmath269 , satisfies the hypothesis of our main theorem .
1em not all symplectic manifolds have a complex structure or a prequantum bundle .
however a symplectic toric manifold , that is when @xmath270 is the dimension of @xmath3 , always admits a compatible complex structure , which is not unique .
furthermore a symplectic toric manifold @xmath3 with momentum map @xmath271 is prequantizable if and only if there exists @xmath272 such that the vertices of the polytope @xmath273 belong to @xmath274 . if it is the case , the prequantum bundle is unique up to isomorphism .
figure [ fig : convex_hull ] shows the joint spectrum of a 4-dimensional toric manifold ( a hirzebruch surface ) . by the atiyah and guillemin - sternberg theorem @xcite , for any hamiltonian torus action on a connected closed manifold ,
the image of the momentum map is a rational convex polytope @xcite .
so the map @xmath275 satisfies the assumption of corollary [ cor : im ] .
even more , for a symplectic toric manifold , the momentum polytope @xmath276 has the additional property that for each vertex @xmath19 of @xmath277 , the primitive normal vectors to the facets meeting at @xmath19 form a basis of the integral lattice @xmath278 .
we call such a polytope a _ delzant polytope_. two symplectic toric manifolds @xmath279 and @xmath280 are _ isomorphic _ if there exists a symplectomorphism @xmath281 such that @xmath282 by the delzant classification theorem @xcite , a symplectic toric manifold is determined up to isomorphism by its momentum polytope .
furthemore , for any delzant polytope @xmath277 , delzant constructed in @xcite a symplectic toric manifold @xmath283 with momentum polytope @xmath277 .
[ theo : inverse - spectral ] let @xmath284 be commuting self - adjoint toeplitz operators on a symplectic toric manifold @xmath285 whose principal symbols are the components of @xmath215 .
then @xmath286 is the delzant polytope @xmath287 .
moreover , @xmath288 is isomorphic with @xmath289 .
the approach of the present paper bypasses the precise description of the semiclassical spectral theory , so it is less informative than the one of @xcite .
however , it has the advantage of concluding isospectrality with an easier proof , which moreover applies in a much more general setting . here
we present an example of a non - toric integrable system modeling a pair of coupled angular momenta as a prequantum bundle of @xmath290 .
it has been described first by d.a .
sadovski and b.i .
zhilinski in @xcite ( see also ( * ? ? ?
* example 6.2 ) and @xcite for further discussion ) .
in order to present this system we need some preliminaries .
consider the unit sphere @xmath291 equipped with the standard area form @xmath292 of total area @xmath293 .
let @xmath294 be the euclidean coordinates on @xmath295 considered as functions on @xmath296 .
the poisson brackets of these functions satisfy the relation @xmath297 and its cyclic permutations .
identify @xmath296 with the complex projective line @xmath298 by the map @xmath299 let @xmath300 be the holomorphic line bundle over @xmath301 dual to the tautological one .
we fix the scheme @xmath119 of the berezin - toeplitz quantization associated to @xmath300 considered as a prequantum bundle of @xmath290 . by a straightforward but cumbersome calculation with berezin s coherent states @xcite
one can verify the quantum commutation relation @xmath302 = -\frac{2i}{m+2 } t_m(z)\ ] ] and its cyclic permutations .
for a positive number @xmath35 , the sphere @xmath303 serves as the phase space of classical angular momentum whose components are given by @xmath304 . the number @xmath35 plays the role of the amplitude of the angular momentum . in view
of we have the relation @xmath305 and its cyclic permutations , where @xmath306 stands for the poisson bracket associated to @xmath307 . if @xmath35 is a positive half - integer , that is @xmath308 , the sphere @xmath303 is quantizable with the prequantum bundle @xmath309 .
the corresponding berezin - toeplitz quantization @xmath310 is given by @xmath311 fix now @xmath312 .
the phase space of the system of coupled angular momenta is the manifold @xmath313 equipped with the symplectic form @xmath314 , while the coupling hamiltonian is independent on @xmath315 and is given by @xmath316 here and below we equip all the data corresponding to the first and the second factor of @xmath3 by lower indices @xmath141 and @xmath174 ( e.g. @xmath317 is the @xmath318-coordinate on the second factor , etc . )
we write @xmath319 for the poisson bracket on @xmath3 .
the coupling hamiltonian @xmath320 admits a first integral @xmath321 .
indeed , by using one readily checks that @xmath322 . in order to quantize this system ,
introduce the prequantum bundle @xmath323 over @xmath3 , where @xmath324 is a copy of @xmath300 over the @xmath14-th factor of @xmath3 , and @xmath325 is the projection of @xmath3 to the @xmath14-th factor , @xmath326 .
denote by @xmath327 the corresponding berezin - toeplitz quantization . for @xmath326
put @xmath328 , and set @xmath329 by and , the operators @xmath330 and @xmath331 are toeplitz operators with the principal symbols @xmath320 and @xmath332 respectively .
the commutation relations readily yield that @xmath333=0 $ ] .
let us emphasize that for a given integrable system @xmath334 , the existence of pair - wise commuting semiclassical operators with principal symbols @xmath334 is not at all automatic , see @xcite for a discussion and references .
next , let us describe the classical spectrum of the system , that is the image of the momentum map @xmath335 without loss of generality assume that @xmath336 and @xmath337 ( otherwise make a rescaling , maybe after the permutation of the variables ) .
it is not hard to see that the image of the map @xmath338 is the spherical shell @xmath339 observe that @xmath340 and @xmath332 is simply the @xmath341-coordinate of @xmath342 .
therefore on each sphere of radius @xmath343\;,\ ] ] in @xmath344 centered at the origin , the value of @xmath332 runs from @xmath345 to @xmath346 .
furthermore , @xmath320 is an increasing function of @xmath346 which takes values @xmath347 at @xmath348 .
we conclude that the image of @xmath349 is the domain @xmath350 given by the inequalities @xmath351 this domain is clearly convex .
we conclude by theorem [ theo : toeplitz ] that _ the convex hull of the joint spectrum of the toeplitz operators @xmath352 and @xmath353 converges to @xmath277 in the hausdorff sense as @xmath354 . _
if @xmath355 or @xmath3 is the cotangent bundle of a closed manifold , then a well - known semiclassical quantization of @xmath3 is given by @xmath1-pseudodifferential operators , which is a semiclassical variant of the standard homogeneous pseudodifferential operators ( see for instance @xcite ) . in this setting , commuting semiclassical pseudodifferential operators have been considered by charbonnel @xcite ; see also @xcite . in the remaining of this text , we omit the @xmath356 prefix for notational simplicity .
symbolic calculus of pseudodifferential operators is known to hold when the symbols belong to a hrmander class .
for instance one can take @xmath357 here @xmath358 for @xmath359 .
if @xmath72 , its weyl quantization is defined on @xmath360 by @xmath361 let @xmath34 be a closed @xmath362-dimensional manifold equipped with a smooth density @xmath215 .
we cover it by a finite set of coordinate charts @xmath363 each of which is identified with a convex bounded domain of @xmath0 equipped with the lebesgue measure ( the existence of such an atlas readily follows from moser s argument @xcite ) .
let @xmath364 be a partition of unity subordinated to @xmath363 , that is @xmath365 and @xmath366 then , for any @xmath367 such that @xmath368 for some @xmath369 , we define , for @xmath370 , @xmath371 where @xmath372 is the weyl quantization calculated in @xmath373 . the operator @xmath374 is a pseudodifferential operator on @xmath34 with principal symbol @xmath375 for @xmath376 .
therefore @xmath77 is a pseudodifferential operator on @xmath34 with principal symbol @xmath377 . the standard pseudodifferential symbolic calculus @xcite gives the following proposition .
@xmath1-pseudodifferential quantization on @xmath34 , where @xmath34 is either @xmath0 or a closed manifold equipped with a density , is a semiclassical quantization of @xmath378 in the sense of definition [ bt ] , where @xmath57 is a hrmander symbol class , @xmath379 $ ] , the hilbert space @xmath380 is @xmath381 ( it is independent of @xmath1 ) . if @xmath382 , @xmath77 is given by the weyl quantization .
if @xmath34 is a closed manifold , @xmath77 is constructed via formula . for the reader
s convenience , let us outline the proof of the proposition .
let @xmath383 be the operator of multiplication by @xmath384 .
then reads @xmath385 , and so @xmath77 is self - adjoint since @xmath386 and the weyl quantization are self - adjoint .
axiom ( q[item : one ] ) follows from the fact that @xmath387 , and @xmath388 .
axiom ( q[item : garding ] ) is known as the grding inequality ( * ? ? ?
* theorem 7.12 ) . axiom ( q[item : symbolic ] ) is a consequence of the product formula for the weyl quantization ( see e.g. ( * ? ? ?
* theorem 7.9 ) ) . for property ( q[item : norm ] ) see ( * ? ? ?
* theorem 13.13 ) .
alternatively , it is not hard to derive it from a standard result on spectral asymptotics for @xmath1-pseudo - differential operators ( see e.g. ( * ? ? ?
* corollary 9.7 ) ) . for reader
s convenience , we present a direct short argument .
let @xmath389 be a pseudo - differential operator with a compactly supported principal symbol @xmath390 .
take @xmath391 such that @xmath392 .
let @xmath100 be a smooth function on @xmath34 supported in a small neigbourhood of @xmath393 , contained in a coordinate chart @xmath394 for @xmath34 , and such that @xmath395 . in these coordinates ,
we define the wkb function on @xmath394 @xmath396 then we may compute directly @xmath397 the phase @xmath398 is stationary when @xmath399 and @xmath400 .
the hessian @xmath401 is non - degenerate .
thus , for fixed @xmath318 in a neighborhood @xmath402 of @xmath393 the stationary phase approximation gives @xmath403 hence @xmath404 furthermore , the remainder in is of order @xmath405 uniformly in @xmath406 .
therefore , shrinking the neighborhood @xmath402 if necessary , we get that @xmath407 for @xmath408 small enough .
this gives the desired result , since @xmath409 this completes the proof .
[ remark - hbar - dependence ] in theorem [ theo : pseudodifferential](ii ) and corollary [ cor : im0 ] the assumption on mild dependence of the symbol on @xmath1 can not be dropped .
indeed , consider a single @xmath1-pseudodifferential operator @xmath410 on @xmath216 . here
the dependence of @xmath411 on @xmath1 is not mild .
the principal symbol vanishes so the classical spectrum equals @xmath412 , while for each @xmath413 the spectrum of @xmath414 equals @xmath216 .
we conclude that the classical spectrum can not be recovered from the quantum one in the classical limit .
let us mention that the formulation and the proof of theorem [ theo : general ] remain valid if we allow a slightly more general class of symbols @xmath36 where all @xmath37 are uniformly ( in @xmath1 ) bounded but not necessarily compactly supported .
the significance of such a generalization is questionable since such symbols do not form an algebra .
it contains however two subspaces that are algebras : @xmath106 as defined in the present paper , and @xmath415 which is defined by requesting @xmath38 to be bounded and @xmath416 to be uniformly bounded .
consider a particle on the plane @xmath417 with potential energy @xmath418 , where @xmath402 is a smooth function on @xmath419 such that @xmath420 , @xmath421 and @xmath422 on @xmath423 .
the hamiltonian of the particle is @xmath424 the corresponding hamiltonian system admits a first integral , the angular momentum @xmath425 .
fix a hrmander class @xmath105 and assume that it contains @xmath320 . a standard calculation with the weyl quantization
shows that @xmath426 the operators @xmath427 commute .
[ lem - leg ] let @xmath428 be the legendre transform of @xmath429
. then @xmath430 , @xmath431 on @xmath423 , and @xmath432 is concave on @xmath433 .
it is immediate that @xmath430 and @xmath431 on @xmath423 .
now assume @xmath434 .
the maximum of @xmath435 is attained at a unique point , say @xmath436 . by a property of the legendre transform , @xmath437 and hence @xmath438
observe also that @xmath439 : indeed , @xmath67 is strictly convex since @xmath440 is strictly convex on @xmath423 .
further , @xmath441 differentiating by @xmath341 we get that @xmath442 where the last inequality follows from @xmath422 . by , @xmath443 in order to prove concavity of @xmath432 it suffices to check that @xmath444 , that is @xmath445 . by , and this is equivalent to @xmath446 , but this follows from equation .
[ prop - rot ] assume that @xmath420 and @xmath421 and @xmath422 on @xmath423 .
then the classical spectrum of @xmath447 is convex , and hence by corollary [ cor : im0](ii ) it can be recovered from the joint spectrum of @xmath448 in the classical limit . put @xmath449 and @xmath450 .
fix an energy level @xmath451 , that is @xmath452 .
let @xmath453 be the maximal possible value of @xmath454 on this level , that is @xmath67 is given by .
let @xmath436 be as in the lemma above . since @xmath455 the value of @xmath456 runs between @xmath457 and @xmath458 .
it follows that the classical spectrum is the subset of @xmath459 defined by @xmath460 , which , by lemma [ lem - leg ] , a is convex set .
let @xmath461 be a separable hilbert space , and let @xmath25 be possibly unbounded selfadjoint operators on @xmath461 . denote by @xmath462 the set of all _ mixed quantum states _ for @xmath25 : the elements of @xmath462 are trace class selfadjoint operators @xmath463 such that here @xmath468 stands for the schatten trace class , and the image of @xmath469 is assumed to lie in the domain of @xmath17 , @xmath470 .
of course @xmath462 is not a vector space , but it is a convex set of compact operators .
the third condition is void in case the operators @xmath17 s are bounded : indeed , the hlder inequality gives @xmath471 ( see also ( * ? ? ?
* theorem vi.19 ) ) .
an interesting subset of @xmath462 consists of the so - called _ pure states _ , i.e. rank - one orthogonal projectors .
we shall call the closure of the image of pure states by @xmath474 the _ joint numerical range _
@xmath475 :
@xmath476 let @xmath478 , and let @xmath479 be a hilbert basis of eigenvectors of @xmath469 so that @xmath480 .
for any @xmath467 , we can write @xmath481 where @xmath482 and @xmath483 . therefore any element of @xmath484 is of the form @xmath485 , where @xmath486 being an infinite convex combination of elements of @xmath475 , @xmath487 must lie in the closed convex hull of @xmath475 , which is equal to the convex hull of @xmath475 since @xmath475 is closed , by definition .
thus @xmath488 the other inclusion follows from @xmath489 and the fact that , since @xmath462 is a convex set and @xmath474 is an affine map , @xmath490 is convex . using the notation of section [ sec : convexity ] , we consider the map @xmath493 , @xmath202 .
notice that @xmath494 where @xmath495 . from lemma [ lem : btcommuting ] , we have @xmath496 therefore , by lemma [ lem : duality ] , @xmath497 the result now follows from lemma [ lemm : numerical - range ] .
recall from and that if @xmath502 , @xmath503 now we repeat the proof of theorem [ theo : general ] .
let @xmath332 be the map of principal symbols of @xmath25 , and write @xmath504 .
by lemma [ lem : max ] , we have that @xmath505 which , in view of , reads @xmath506 to finish the proof we repeat the proof of theorem [ theo : bounded ] , where the convex hull @xmath492 is replaced by @xmath491 , and ( [ equ : pointwise ] ) by . _ acknowledgements .
_ ap was partially supported by nsf grants dms-0965738 and dms-0635607 , an nsf career award , a leibniz fellowship , spanish ministry of science grants mtm 2010 - 21186-c02 - 01 and sev-2011 - 0087 .
lp is partially supported by the national science foundation grant dms-1006610 and the israel science foundation grant 509/07 .
he gratefully acknowledges the warm hospitality from the university of chicago where part of this project has been fulfilled .
vns is partially supported by the institut universitaire de france , the lebesgue center ( anr labex lebesgue ) , and the anr nosevol grant .
he gratefully acknowledges the hospitality of the ias .
s. v ngc : symplectic inverse spectral theory for pseudodifferential operators .
geometric aspects of analysis and mechanics _ progress in mathematics _
birkhuser / springer , new york , 2011 , volume 292 , 353 - 372 .
+ * lvaro pelayo * + school of mathematics + institute for advanced study + einstein drive + princeton , nj 08540 usa .
+ + washington university , mathematics department + one brookings drive , campus box 1146 + st louis , mo 63130 - 4899 , usa . + _ e - mail _ : ` [email protected] ` + _ website _ : http://www.math.wustl.edu/~apelayo/ * san v ngoc * + institut universitaire de france + + institut de recherches mathmatiques de rennes + universit de rennes 1 + campus de beaulieu + f-35042 rennes cedex , france + _ e - mail : _ ` [email protected] ` + _ website _ : http://blogperso.univ-rennes1.fr/san.vu-ngoc/ | we introduce a minimalistic notion of semiclassical quantization and use it to prove that the convex hull of the semiclassical spectrum of a quantum system given by a collection of commuting operators converges to the convex hull of the spectrum of the associated classical system .
this gives a quick alternative solution to the isospectrality problem for quantum toric systems . if the operators are uniformly bounded , the convergence is uniform .
analogous results hold for non - commuting operators . |
weak semileptonic decays of charm mesons continue to attract interest due to the relative simplicity of their theoretical description : the matrix element of these decays can be factorized as the product of the leptonic and hadronic currents .
this makes the @xmath0 decay a natural place to study the @xmath11 system in the absence of interactions with other hadrons .
it is known that the @xmath11 final state of @xmath0 decay is strongly dominated by the @xmath4 vector resonance @xcite .
the large and clean sample of @xmath12 events collected by the fermilab focus experiment provides an excellent opportunity to measure the @xmath4 mass and width , as well as the effective blatt - weisskopf damping factor parameter discussed in ref .
@xcite .
in addition we also search for structures other than the @xmath4 resonance in the mass range of 0.65 gev/@xmath1 @xmath13 1.5 gev/@xmath1 .
the available phase space allows us to investigate the low mass scalar sector whose spectroscopy has been increasingly studied since the indication of the existence of the broad scalar resonance @xmath14 in @xmath15 decays from a dalitz plot analysis performed by e791 @xcite .
recently e791 has also pursued an independent approach for the scalar component extracting directly the scalar phase shift from their data @xcite .
these two results , obtained using @xmath15 hadronic decay is substantially different from the one obtained in a partial wave analysis performed by the lass scattering experiment @xcite .
lass observed that the s - wave amplitude in the @xmath16 elastic region can be represented as the sum of a @xmath17 resonance and a smooth shape , consistent with the non - resonant hypothesis .
figure [ lass_kpipi ] shows a comparison of the phase shift of the scalar component found by lass to the one found by e791 in @xmath15 decay .
scattering phase shift measured by lass and the one measured by the e791 experiment ( solid circles ) in @xmath15 charm hadronic decay ( figure provided by b. meadows ) .
[ lass_kpipi],width=340,height=226 ] due to watson s theorem @xcite the observed @xmath11 phase shifts in @xmath0 decays should be the same as those measured in @xmath11 elastic scattering .
this establishes a connection between the production and the scattering mechanisms , adding crucial information about @xmath11 hadronic interaction and the role of the bachelor @xmath18 final state interaction in hadronic decays .
the data were collected in the wideband photoproduction experiment focus during the fermilab 19961997 fixed - target run . in focus , a forward multi - particle spectrometer is used to measure the interactions of high energy photons on a segmented beo target .
the focus detector is a large aperture , fixed - target spectrometer with excellent vertexing and particle identification .
the focus beamline @xcite , the detector @xcite and the specific cut selection @xcite have been described elsewhere .
a total of 18245 @xmath0 candidates remain after the selection criteria .
we estimate the charm and non - charm background contributions to be , respectively , 17.8% and 3.2% of the total number of events over our signal region .
four - body decays of spinless particles are described by five kinematic variables .
the variables chosen in this analysis are the @xmath16 invariant mass ( @xmath19 ) , the square of the @xmath20 mass ( @xmath21 ) , and three decay angles : the angle between the @xmath22 and the @xmath23 direction in the @xmath16 rest frame ( @xmath24 ) , which defines one decay plane , the angle between the @xmath25 and the @xmath23 direction in the @xmath20 rest frame ( @xmath26 ) , which defines the second decay plane , and the acoplanarity angle ( @xmath27 ) between these two decay planes .
the differential decay rate can be represented by a coherent sum of resonant and non - resonant contributions to the angular momentum eigenstates of the @xmath16 system , @xmath28 where @xmath29 , @xmath30 is the weak matrix element for a transition with angular momentum @xmath31 , @xmath32 represents the form of the hadronic final state amplitude contribution of resonance @xmath33 ( or non - resonant ) with strength @xmath34 , and @xmath35 is the phase space density .
the possible resonant states that couple to @xmath16 are the scalars @xmath14 and @xmath17 , the vectors @xmath4 and @xmath36 .
mass spectrum .
the inclusion of a small @xmath37 resonance contribution is unlikely to be observed , since it is orthogonal to the ( dominant ) @xmath4 and low mass s - wave amplitudes . ]
the non - resonant contribution is assumed to be scalar .
the parametrization of resonant states with angular momentum @xmath31 is given by the product of a breit - wigner and the normalized @xmath38 coupling , @xmath39 @xmath40 where @xmath41 , @xmath42 is the magnitude of the kaon momentum in the resonance rest frame , @xmath43 , @xmath44 , and @xmath45 .
@xmath46 is the blatt - weisskopf damping factor given by @xmath47 @xcite .
the damping factor adds an additional fit parameter , @xmath48 , in our fits to the @xmath4 line shape .
the line shape of the @xmath14 resonance is expected to deviate significantly from a pure breit - wigner , due to its large width and the close vicinity of the @xmath11 threshold . in this analysis
we use the @xmath14 line - shape adopted by e791 @xcite while for the non - resonant component we use an empirical phase shift , parametrized by the effective range model , obtained by lass ( fig .
[ lass_kpipi ] ) . , from lass non - resonant amplitude , which is already included in eq .
[ eq1 ] . ] the weak matrix element for the vector process , @xmath49 , and for the scalar process , @xmath50 , are written as a function of helicity amplitudes , @xmath51 , derived in @xcite .
next we discuss the angular distribution described by eq .
the @xmath11 spectrum described by this equation includes the dominant contribution from the @xmath4 resonance , possible high mass contributions from the @xmath17 and @xmath36 resonances , and low mass scalar components comprised of a non - resonant and a possible @xmath14 contributions , both populating the region where significant discrepancies were found between the data and the predicted @xmath52 angular decay distributions .
it has been shown in @xcite that a nearly constant amplitude and phase contribution to the helicity zero amplitude of the virtual @xmath53 was required to adequately fit the observed decay angular distributions .
the @xmath19 distribution weighted by @xmath54 provides information on the phase of the additional structure relative to that of the @xmath4 .
it can be used to discriminate different combinations of low mass states , given the large difference between their phase shifts .
figure [ fig_asymmetry ] compares the distribution obtained in the data with the predictions from the non - resonant and @xmath14 models in the absence of additional phase shifts as required by watson s theorem . since a simulation using the lass parametrization of the non - resonant contribution is sufficient to reproduce the data , we exclude a possible @xmath14 contribution from further consideration . weighted by @xmath54 .
the data ( squares ) show good agreement with the lass non - resonant parameterization ( solid histogram ) but not with a @xmath14 model ( dashed histogram),width=264,height=188 ]
having excluded the @xmath14 , the most general differential decay rate for @xmath0 in @xmath19 is given by eq .
[ eq5b ] .
@xmath55 the @xmath56 , are real functions will appear as sinusoidal functions of @xmath27 .
hence any imaginary terms vanish when averaged over @xmath27 given our nearly uniform acceptance in this variable . ] that depend only on @xmath19 .
these functions are computed from the @xmath19 spectrum obtained from a complete simulation of @xmath0 events , generated according to phase space and weighted by @xmath57 and thus represent the intensity modified by acceptance and efficiency , @xmath58 . the vector , @xmath59 , and scalar , @xmath60 , amplitudes are given by the sum of the relevant states , each one represented by the product of a real magnitude , @xmath61 , and @xmath62 .
the @xmath63 , @xmath64 , and @xmath65 functions depend on @xmath19 as well as on all fit parameters .
the cross - term , @xmath66 , represents the interference between the vector and scalar contributions .
the fit parameters are the magnitudes of each amplitude , @xmath67 , the mass and width of the @xmath4 , and the parameter @xmath48 of the blatt - weisskopf damping factor .
the parameters of all other resonances are fixed to the pdg values @xcite .
the fit parameters are obtained from an unbinned maximum - likelihood method , by minimizing the quantity @xmath68}$ ] , where @xmath69 is the probability density for the signal , @xmath70 , for the background , and @xmath71 is the background fraction , fixed in the estimated values .
the @xmath4 amplitude is taken as the reference amplitude .
decay fractions are obtained integrating each individual amplitude over the phase space and dividing by the integral over the phase space of the overall amplitude . to account for momentum resolution effects on the @xmath4 parameters , we refit the data fixing all parameters except the @xmath4 width and use a probability density function convoluted with a gaussian distribution with @xmath72 , value obtained from monte carlo simulation .
the smearing due to momentum resolution increases the @xmath4 width by approximately @xmath73 .
using the procedure described above , we fit the data assuming only a @xmath52 process .
the confidence level of this fit is 0.21% , indicating the need for additional contributions in the decay .
the inclusion of a non - resonant scalar component , referred to as the _ nr model _ , significantly improves the confidence level of the fit to 66% .
we find @xmath74 mev / c@xmath75 , @xmath76 mev / c@xmath75 , @xmath77 gev@xmath8 and @xmath78 , which correspond to a scalar fraction of @xmath79 % .
figure [ fig_results ] illustrates the contribution of both the @xmath52 and non - resonant s - wave process to the observed @xmath19 spectrum .
data using the nr model .
the error bars , the solid lines , the dashed lines , and the dotted lines correspond to the data , the model , the background contribution , and the scalar contribution , respectively .
the upper right plot shows the same information and the cross - term ( dot - dash line ) with a limited y - axis to allow more detail to be seen .
[ fig_results],width=264,height=226 ] we also consider @xmath80 and @xmath81 contributions to our model
. the observed @xmath36 and @xmath17 amplitudes are consistent with zero and we find @xmath82% and @xmath83% at 90% cl .
the upper limits are calculated using the method described in @xcite and assume @xmath84 and @xmath85 @xcite . to study the statistical significance of these new amplitudes
, we use a hypothesis test based on the maximum - likelihood ratio method @xcite . as a result
, we obtain a confidence level of 80% in favor of the simple nr model .
in conclusion we find that our angular distribution is consistent with the effective - range scalar non - resonant phase shift obtained by lass @xcite as expected by watson s theorem given the absence of other final state interactions ( fsi ) .
it points out that the fsi of the bachelor pion of the @xmath86 decay can play an important role towards the observed difference with respect to @xmath11 scalar scattering phase shift .
we have measured the @xmath4 parameters using a large sample of @xmath52 signal events over a wide mass range . the absence of high mass resonances as well as the small background contribution provides a unique environment to study the @xmath4 mass and width .
our measurements of the mass and width are more than @xmath87 below the present world average value .
we obtain a blatt - weisskopf parameter consistent with the value obtained by lass @xcite .
we also limit possible additional @xmath11 resonances present in @xmath0 semileptonic decays .
we wish to acknowledge the assistance of the staffs of fermi national accelerator laboratory , the university of rome `` tor vergata '' , the infn of italy , the universidade federal do rio de janeiro , the fundao de amparo pesquisa do estado do rio de janeiro and the physics departments of the collaborating institutions .
this research was supported in part by the u. s. national science foundation , the u. s. department of energy , the italian instituto nazionale di fisica nucleare and ministero delluniversit e della ricerca scientifica e tecnolgica , the brazilian conselho nacional de desenvolvimento cientfico e tecnolgico , conacyt - mxico , the korean ministry of education , and the korea research foundation . | we present a four - body semileptonic charm decay @xmath0 analysis in the range of 0.65 gev/@xmath1 @xmath2 1.5 gev/@xmath1 .
we observe a low mass scalar contribution of @xmath3 with respect to the total @xmath0 decay , compatible with the phase shift found by the lass elastic scattering experiment . for the @xmath4 resonance ,
we obtain a mass of @xmath5 mev/@xmath1 , a width of @xmath6 mev/@xmath1 , and a blatt - weisskopf damping factor parameter of @xmath7 gev@xmath8 . we also report 90% cl upper limits of 4% and 0.64% for the branching ratios
@xmath9 and @xmath10 , respectively . |
recent experimental observations of macroscopic quantum tunneling ( mqt ) @xcite and energy level quantization @xcite in high-@xmath0 superconductor josephson junctions ( jjs ) open up the possibility for realizing high-@xmath0 quantum bits . in ybco grain boundary junctions which was used in mqt experiments @xcite , it was found that the stray capacitance @xmath1 of the electrodes , due to the large dielectric constant of the sto substrate at low temperature ( @xmath2 ) , and the inductance @xmath3 , due to the large london penetration depth in @xmath4-axis and/or josephson coupling between cuo@xmath5 planes in one of the electrodes have large influence on the macroscopic dynamics and can be taken into account by an extended circuit model ( see fig.1(a ) ) . in fig . 1 , @xmath6 is the phase difference across the josephson junction and @xmath7 is the phase difference across the capacitor @xmath1 , where @xmath8 is the current through the inductor .
it was found that , in the microwave - assisted mqt experiment @xcite , the bias current @xmath9 dependence of the josephson plasma frequency @xmath10 is quantitatively explained by this model @xcite .
however , the validity of the extended circuit model for ta and mqt escape processes have not yet been explored .
below we calculate the mqt and ta escape rate base on the extended circuit model and try to compare with experimental results . for @xmath11 and @xmath12.,width=453 ]
the hamiltonian of the circuit ( fig .
1(a ) ) can be written as @xmath13 where @xmath14 , @xmath15 , @xmath16 is the josephson coupling energy , @xmath17 , and @xmath18 is the external current ) .
this hamiltonian describes the quantum dynamics of a fictive particle moving in the 2-dimensional tilted washboard potential @xmath19 ( see fig.1(b ) ) . by introducing the new
coordinate @xmath20 and assuming @xmath21 , we can rewrite the hamiltonian as @xmath22 where @xmath23 is the local minimum point , @xmath24 , and @xmath25 . by using the functional integral method @xcite ,
the partition function of the system can be written as @xmath26 / \hbar \right ] , $ ] where @xmath27 $ ] is the euclidean lagrangian .
the lagrangian is a quadratic function of @xmath28 .
therefore the functional integral over variable @xmath28 can be performed explicitly .
then the partition function is reduced to a single functional integral over @xmath29 , @xmath30 , @xmath31 /\hbar)$ ] , where the effective action is given by @xmath32 & = & \int_0^{\hbar \beta } d \tau \left [ \frac{1}{2 } m \dot{x}^2 + u_\mathrm{1d } ( x ) \right ] + \frac{1}{4 } \int_{0}^{\hbar \beta } d \tau \int_{0}^{\hbar \beta } d \tau ' \left [ x(\tau ) - x(\tau ' ) \right]^2 k ( \tau-\tau ' ) .\end{aligned}\ ] ] in this equation , @xmath33 $ ] , and @xmath34 } { \sinh \left [ \frac { \hbar \beta \omega_{lc } } { 2 } \right ] } , \end{aligned}\ ] ] where @xmath35 is the resonance frequency of the lc circuit .
thus the dynamics of the phase difference in the 2d potential @xmath19 can be mapped into simple 1d model .
note that , due to the coupling between the josephson junction and the lc circuit , @xmath36 $ ] contains a `` dissipation '' term ( the second term ) which is nonlocal in imaginary time @xmath37 .
the thermal activation rate well above the crossover temperature is given by @xcite @xmath38 where the effective trial frequency @xmath39 is a solution of @xmath40 , @xmath41 is the matsubara frequency , @xmath42 is the josephson plasma frequency , and @xmath43 is the damping function .
note that the potential barrier height @xmath44 ( @xmath45 ) is not modified even in the presence of the lc circuit .
therefore the coupling to the lc circuit only modifies the prefactor of @xmath46 .
the effective trial frequency @xmath39 and the quantum correction @xmath47 in the prefactor can be calculated analytically , @xmath48 .
, @xmath49 for the non non - adiabatic cases ( @xmath50 ) which is applicable to the mqt experiment in ybco grain - boundary jjs @xcite . in the case of @xmath51 ,
the prefactor of @xmath46 coincides with the result without dissipative effects @xcite .
therefore we can conclude that the influence of the coupling to the lc circuit on the thermal activation process is quite weak .
so the system behaves as simple 1d systems well above the crossover temperature .
this result is qualitatively consistent with the experimental observation @xcite .
the mqt escape rate at zero temperature is defined by @xmath52 @xcite . in order to determine @xmath53
, we use the bounce techniques .
then we get @xmath54 where @xmath55 + { \cal s}_\mathrm{diss } [ x_b])/\hbar$ ] is the bounce exponent , that is the value of the action @xmath56 evaluated along the bounce trajectory @xmath57 . in the non - adiabatic case ( @xmath58 ) , we get @xmath59 , \end{aligned}\ ] ] where the `` dissipation '' correction term is given by @xmath60 .
in contrast to @xmath46 , the coupling to the lc circuit gives rise to an increase of the bounce exponent .
therefore the two - dimensional nature has large influence on the mqt escape process . in order to check the validity of our model
, we try to compare our result with the experimental data of the switching current distribution at the low temperature regime @xcite .
the switching current distribution @xmath61 is related to the mqt rate @xmath62 as @xmath63 , \end{aligned}\ ] ] where @xmath64 is the sweep rate of the external bias current . from the experiment @xcite
, it is found that the measured value of the full width at half maximum ( hmfw ) @xmath65 of @xmath61 is 11.9na at the temperature - independent mqt regime . on the other hand , in order to numerically calculate @xmath65 , we need the information of @xmath66 , @xmath67 , @xmath3 , and @xmath1 .
the values of @xmath68a , @xmath69nh and @xmath70pf were directly determined from the ta and the microwave - assisted mqt experiments @xcite . therefore the only fitting parameter is @xmath67 . from the numerical estimation of @xmath65
, we found that @xmath71 pf gives good agreement with the experimental value of @xmath65 .
the obtained value of @xmath67 is consistent with the estimated value @xmath72pf based on the geometry of the junction @xcite .
therefore , we can conclude that the extended circuit model can quantitatively explain the mqt experiment in the ybco grain boundary junctions .
we have theoretically studied the ta and mqt escape for the josephson junction coupled with a lc circuit by taking into account the two - dimensional nature of the phase dynamics .
then we found that the coupling to the lc circuit gives negligible reduction for ta escape rate . on the other hand
, we also found that the mqt escape rate is largely reduced due to the coupling with the lc circuit .
these results are consistent with experimental data of ybco josephson junctions .
we would like to thank j. ankerhold , m. fogelstrm , g. johansson , j. r. kirtley , j. p. pekola , g. rotoli , v. shumeiko , and f. tafuri for useful discussions .
this work was supported by jst - crest and the jsps - rsas scientist exchange program .
+ 99 bauch t
_ et al _ 2005 _ phys .
lett . _ * 94 * 087003 inomata k _ et al _ 2005 _ phys .
lett . _ * 95 * 107005 matsumoto t _
et al _ 2007 _ supercond .
technol . _ * 20 * s10 li s x _ et al _ 2007 _ phys .
* 99 * 037002 kashiwaya h _ et al _ 2008 submitted to _ j. phys
et al _ 2006 _ science _ * 311 * 57 jin x y _ et al _ 2006 _ phys . rev .
* 96 * 177003 lombardi f _ et al _ 2007 _ ieee trans .
_ * 17 * 653 rotoli g
_ et al _ 2007 _ phys .
b _ * 75 * 144501 weiss u 2008 _ quantum dissipative systems _
( world scientific , singapore ) kawabata s _ et al _ 2004 _ phys .
b _ * 70 * 132505 kawabata s _ et al _ 2005 _ phys . rev .
b _ * 72 * 052506 kawabata s _ et al _ 2006 _ phys . rev . b _ * 74 * 180502(r ) kawabata s _ et al _ 2007 _ phys .
b _ * 76 * 064505 yokoyama t _
et al _ 2007 _ phys .
b _ * 76 * 134501 kawabata s _ et al _ 2007 _ supercond .
technol . _ * 20 * s6 | we investigate classical thermal activation ( ta ) and macroscopic quantum tunneling ( mqt ) for a josephson junction coupled with an lc circuit theoretically .
the ta and mqt escape rate are calculated analytically by taking into account the two - dimensional nature of the classical and quantum phase dynamics .
we find that the mqt escape rate is largely suppressed by the coupling to the lc circuit . on the other hand ,
this coupling gives rise to slight reduction of the ta escape rate .
these results are relevant for the interpretation of a recent experiment on the mqt and ta phenomena in grain boundary ybco josephson junctions . |
lofar ( the low frequency array ) is a new kind of radio telescope for astronomical observations in the frequency range of @xmath1 mhz with high sensitivity and high spatial resolution ( http://www.lofar.org ) .
it uses a large number of simple dipole antennas instead of the traditional big parabolic dishes .
it consists of @xmath2 stations in the netherlands , @xmath3 in germany and one each in great britain , france and sweden covering a total area of more than @xmath4 km in diameter . though primarily design as an astronomical telescope , lofar can also be used for the detection of very high - energy cosmic rays ( crs ) in the interesting energy region above @xmath5 ev where the transition of galactic to extra - galactic crs is expected ( hrandel et al . 2009 , horneffer et al .
2010 ) . this will be done by looking at extensive air showers which are essentially cascades of energetic secondary particles produce by the interaction of cr primaries with the nuclei present in the atmosphere .
a large fraction of these secondaries are electrons and positrons which produce radio synchrotron emission in the presence of the earth s magnetic field ( falcke et al . 2005 ) .
due to coherence effects , this emission can give strong signals on the ground in the frequency range of @xmath6 mhz detectable by the lofar low band antennas .
in addition , lofar can also look for the highest energy crs or neutrinos above around @xmath7 ev ( scholten 2010 ) .
this will be carried out by detecting coherent radio cherenkov emission from particle cascades at the moon induced by those crs or neutrinos . at frequencies of @xmath8 mhz which is within the range of the lofar high band antennas ,
the angular spread of this emission becomes wider leading to more detectable signals at the earth .
the different energy ranges detectable by lofar are shown in fig . 1 where the vhecr ( very high energy crs ) refers to the detection using air showers and the uhep ( ultra high energy particles ) refers to that using the moon .
1 also shows the energy ranges for the lopes ( lofar prototype station ) experiment which is located at the kascade - grande experimental site ( apel et al .
2010 ) and the currently building aera ( auger engineering radio array ) experiment at the pierre auger observatory ( van den berg et al .
2009 ) . one important goal of the lofar cr experiment is to push the radio detection technique towards an independent way of detecting very high energy crs . by detecting the radio signals with better sensitivity and better spatial resolution in a wider frequency range
, we strongly believe that lofar will provide better understanding of the measured signals , their emission mechanisms and their relations with the air shower parameters , thereby leading to better estimates of the properties of the primary particle .
compared to the lopes ( 30 antennas ) and the codalema ( 24 antennas ) experiments ( ardrouin et al . 2005 ) , lofar has @xmath9 stations in its core ( an area of @xmath10 km@xmath11 ) with each station consisting of @xmath12 low band antennas and @xmath13 high band antennas . however , at this stage , it is still quite early for radio detection experiments to do a stand - alone study on crs . therefore , we are also setting up a small particle detector array called lora ( lofar radboud air shower array )
. its main role will be to trigger and confirm the radio detection of air showers with the lofar antennas .
it will also help in the reconstruction of several important air shower parameters like the primary energy , the shower core location , the arrival direction , the lateral density distribution etc .
lora is setting up inside the lofar core within an area of @xmath0 m diameter .
it consists of 5 stations with 4 particle detectors each , placed at a separation of @xmath14 m between them .
it is expected to detect crs with energies greater than @xmath15 ev at an event rate of around once every few minutes .
detailed simulations about lora as well as combined studies including the lofar radio data will be carried out as soon as possible .
the schematic layout of the lora detectors arrangement is shown in fig . 2 where the red stars represent the positions of the detectors and the circles denote the positions of the lofar antennas fields .
data collection in individual stations are controlled locally by station computers which are then controlled by a master computer where the overall data processing is done .
m diameter with relative spacings of @xmath14 m.,width=264 ] the detectors for lora are taken over from the kascade experiment ( antoni et al .
each detector is of @xmath16 cm in size and consists of two slabs of @xmath17 cm thick plastic scintillators . when a charged particle passes through a scintillator slab , the light which are produced are collected into a photomultiplier tube ( pmt ) through a wavelength shifter bar .
the electrical signals from the pmts ( one pmt for each slab ) are then combined and fed into the electronics where they are converted into digital signals .
the electronics was developed originally for the hisparc experiment ( http://www.hisparc.nl ) . an electronics unit ( hereafter referred to as hisparc unit )
can handle two channels ( one channel corresponds to one detector ) and there are two units per station ( a master and a slave ) .
the master sets the event trigger condition and also provides the time stamp with a gps receiver along with a @xmath18 mhz clock counter .
data in each channel are handled by two @xmath19-bit adcs ( so there are 4 adcs per hisparc unit ) which can measure voltages in the range of @xmath20 v. event data ( signal ) in a channel are sampled with a time resolution of @xmath21 ns and are stored in a total time window of @xmath22 @xmath23s .
3 shows a typical signal of a charged particle passing through one of our detectors .
data from the electronics are sent to the station computer through usb .
each hisparc unit has a fpga circuit built into it .
this gives an observer control over several parameters required by the electronics as well as the detectors .
for instance , one can set the trigger thresholds , high voltages , trigger condition , coincidence time window etc . at any time during the observation through a user software running at the station computer .
all the pmts are gain calibrated by adjusting their voltages to match the response for single muons .
when the signals in the two hisparc units in a station satisfy some minimum trigger condition , the signals are sent to the station computer which then sends them to the master computer .
the master computer combines the signals from all stations and checks for good air shower events .
when an air shower candidate is found , a first - level online analysis is performed to calculate shower parameters like the arrival direction , shower core position , event size etc . , and a trigger information will be sent to the lofar antenna to dump the corresponding radio data for the air shower .
the master computer takes less than @xmath22 ms to process an event .
this is necessary because the antennas have to dump their radio data corresponding to a particular event before they are overwritten in a memory ring - buffer , known as transient buffer board ( tbb ) .
the daq software for lora is developed in c / c++ language for linux based operating systems .
the software ( particularly the online monitoring tool ) uses several features of the root package ( http://root.cern.ch ) . on the monitoring panel ,
important information is displayed which is useful to monitor the performance of the system electronics and of the detectors during observations .
the final data are stored in root format and they consist of four kinds of data .
the first are the event data which are generated whenever an air shower event is detected .
the data contain the event time stamp for each station and the signal trace in adc counts for each detector . the second kind of data stores the so - called one second messages from the hisparc unit .
this data is generated every second by the master device and it contains information about the number of times the analog signal went over the threshold in the last second for each of the four channels .
it also has important timing information which can be used for calculating an event time stamp with nanosecond accuracy .
the third data kind comprises several control parameters for the observation run .
this data is stored every interval of time fixed by the observer at the start of the run .
the fourth kind of data contains information about the noise level in each channel averaged over some fixed interval of time .
the daq software and a preliminary data analysis software have been tested successfully on a lora prototype installed at radboud university ( ru ) nijmegen .
the results from the prototype are presented in the following section . for the 5463 events collected with the lora prototype.,width=347 ] ( top ) and azimuthal angles @xmath24 ( bottom )
distribution for the measured showers .
the @xmath25 distribution is fitted with a function given by eq .
the @xmath24 distribution is fitted with a constant.,title="fig:",width=328 ] + ( top ) and azimuthal angles @xmath24 ( bottom ) distribution for the measured showers .
the @xmath25 distribution is fitted with a function given by eq .
the @xmath24 distribution is fitted with a constant.,title="fig:",width=328 ]
the lora prototype that we set up to test the electronics and the daq software consists of 4 detectors in a @xmath26 m arrangement . with that
, we collected more than @xmath27 air shower events in a total observation time of @xmath28 hrs ( @xmath29 days ) during april @xmath30 .
this corresponds to an event rate of around once every @xmath17 minutes .
we have performed a preliminary analysis of the data to reconstruct the shower core position on the ground , the arrival direction and the energy of the primary . the primary energy is calculated from the total number of charge particles on the ground using the parameterization given in hrandel 2007 .
this simple estimate gives the energy threshold of the prototype to be less than @xmath31 ev .
the shower core position is calculated using the center of gravity method and the primary arrival direction is calculated using the relative pulse timing informations in the detectors assuming that the shower front arrives in a plane on the ground .
the normal to the shower plane gives the arrival direction and it is characterized by two angles : the zenith angle @xmath25 measured from the vertical direction and the azimuthal angle @xmath24 measured in the horizontal plane . though detailed simulation is still yet to be performed , the energy resolution of lora is expected to be @xmath32 and the angular resolution to be @xmath33 . in figs .
4 and 5 , we show the distributions of the reconstructed core positions and the arrival directions @xmath34 respectively . in fig .
6 , we plot the @xmath25 ( top ) and the @xmath24 ( bottom ) distributions separately .
the @xmath25 distribution is fitted using the following function , @xmath35 the fit parameters are found to be @xmath36 and @xmath37 . for the @xmath24 distribution , we fit it to a constant as @xmath38 eq .
( 2 ) is expected if the arrival direction of the crs are isotropic . however ,
in our case we see a sinusoidal distribution with two dips one at @xmath39 and the other at @xmath40 which are due to the narrow arrangement of our detector set - up .
it should be mentioned that at this stage , we do not aim to perform a detailed analysis and derive a final set of fit parameters which determine important properties of the crs .
as already mentioned , our primary aim was to test our electronics and the daq software .
the preliminary results from the prototype are in general consistent with what we expect from cr observations with air shower arrays .
the electronics and the daq software for lora have been tested successfully .
the installation of all the lora stations in the field is already completed and the set - up is currently under testing . on the other hand , the lofar cr pipeline software is under development and we are expecting to start the simultaneous observation of crs with the radio antennas before summer 2011 . + + * acknowledgements * : the authors are deeply thankful to the kascade collaboration for kindly providing us some of the scintillators previously used for their experiment
. the authors are also grateful to both the anonymous referees for constructive comments .
antoni t. , et al .
2003 , nucl .
instrum . meth .
a @xmath41 490 apel , w. d. , et al .
2010 , astropart .
@xmath42 , 294 ardrouin , d. , et al .
2005 , nucl .
instrum . meth .
a @xmath43 , 148 falcke h. , et al .
2005 , nature , @xmath44 313 hrandel , j. r. , 2007 , mod .
a @xmath45 1533 hrandel j. r. , et al .
2009 , nucl .
phys . b @xmath46 289 horneffer a. , et al 2010 , nucl .
instrum . meth . a @xmath47 482 scholten o. , 2010 , j. phys .
@xmath48 012003 van den berg , a. m. , et al .
2009 , proc .
31th icrc , lodz | lofar is a new form of radio telescope which can detect radio emission from air showers induced by very high - energy cosmic rays .
it can also look for radio emission from particle cascades on the moon induced by ultra high - energy cosmic rays or neutrinos . to complement the radio detection
, we are setting up a small particle detector array lora ( lofar radboud air shower array ) within an area of @xmath0 m diameter in the lofar core .
it will help in triggering and confirming the radio detection of air showers with the lofar antennas . in this paper
, we present a short overview about lora and discuss its current status . |
thermonuclear reactions are of enormous importance for the physics of stars and of our sun .
the general formalism that describes these reactions exists since a long time and there is a wide consensus about our good understanding of the relevant physics @xcite .
however , quantitative calculations of specific reaction rates need experimental inputs and theoretical assumptions .
_ cross sections must be known_. in the best cases , they can be directly measured at the relevant energies ; other cross sections can be only measured at higher energies and need to be extrapolated ; in some situations only theoretical predictions exist ( for a recent review of the cross sections relevant to the sun see ref .
in addition , rates depend on weighted thermal averages of the cross sections ; therefore , knowledge about _ the thermal distribution is also needed_. because many nuclear reactions in the stellar burning core proceed by way of quantum penetration of a high coulomb barrier , their cross sections grow exponentially with energy . therefore , thermal averages do not probe the average energy of the distribution ( @xmath2 ) , but its high - energy tail .
consequently , _ rates are sensitive to a relatively small part of the distribution_. the ion velocity distribution in stellar calculations is always assumed maxwellian .
in fact , it is almost universally accepted as a fact that physical conditions in the solar interior ( density and temperature ) lead to an equilibrium velocity distribution that is maxwellian . in this article
we review the assumptions underlying this general statement , discuss why they are only approximately true , and argue that the consequent _ small corrections to the distribution have significant effects on the rates_. the dynamics of the solar plasma is not trivial , since at such densities and temperatures there is no clear scale separation between collective and individual degrees of freedom @xcite .
the presence of more than just one energy scale ( @xmath2 ) in the relevant range of energies results in deviations from the pure exponential behavior , @xmath3 , which is determined only by @xmath2 .
we have tried several new approaches to this strongly interacting many - body system for which a realistic microscopic calculation does not exist yet . in one approach @xcite
we exploit the knowledge that the distribution in the solar interior can not be too much different from the maxwellian one and add small corrections ( higher - order terms in a derivative expansion ) to the coefficients of the standard fokker - plank equation .
tsallis @xcite and druyvenstein - like distributions are immediately generated . a second approach @xcite focuses on the electric microfields that have been shown to exists in plasmas and tries to link their distribution and the connected effective cross sections with deviations from the maxwellian distribution .
different classes of microfields distributions and effective cross sections yield several non - maxwellian distributions among which we again find tsallis and druyvenstein - like distributions .
the third approach has been just started and aims to connect the distribution of collective variables @xcite to memory effects and long - time correlations between velocities .
there should exist solutions compatible with the tsallis distribution and/or other non - maxwellian distributions .
these three approaches are not exhaustive and not necessary alternative .
nevertheless , it is suggestive that all of them point in the same direction : _ the maxwell - boltzmann distribution of velocity should have small but nonnegligible corrections in the solar plasma and the tsallis distribution could provide a better description_. nowadays solar modeling seems to have reached a satisfactory stage @xcite . the inclusion in the latest models of higher - order effects , such as the diffusion of heavy elements , brings the theoretical predictions in good agreement even with the detailed helioseismological data @xcite .
however , there still exist a discrepancy between the solar neutrino experiments and the predicted neutrino fluxes @xcite . in this context
, there has been a considerable amount of work devoted to answering questions such as : how large are the uncertainties of the solar model input parameters ?
has something been left out of standard solar models ?
how does this affects predictions for the fluxes and the status of solar neutrino problem ( snp ) @xcite ?
therefore , we find extremely important to assess the consequences of the possible small deviations from the standard statistics on the solar model and , in particular , on the neutrino fluxes .
at least two effects should be considerd : changes of the total rates caused by the different thermal average and modifications of the shape of the energy spectra . the effect on the rates
was already considered by clayton @xcite two decades ago .
new experimental data , better solar models and a considerable better understanding of nonstandard distributions and of how they can arise in solar plasmas convinced us of the necessity of reconsider this earlier suggestion . even if experimental detection of small modifications of the neutrino spectra might appear still far from being possible , it is useful to have a clear assessment of what the signal would be .
our paper is organized as follows . in sec .
[ secrate ] we review the physics of subbarrier thermonuclear reactions and how the velocity distribution influences reaction rates . section [ secplasma ] , which is dedicated to the solar plasma , contains the central part of our work : three approaches to velocity distribution calculation and the resulting nonstandard distributions . in sec .
[ secmodirate ] we calculate the reaction rates with the modified distributions , while we obtain the neutrino spectrum from the @xmath1 reaction in the presence of tsallis statistics and compare it to the standard one in sec .
[ secspectrum ] .
the solar neutrino problem is analysed in the light of our results on nonstandard velocity distributions in sec .
[ secsnp ] , and sec .
[ secconclu ] is reserved to our conclusions .
reliable calculations of nuclear reaction rates in stellar interiors is fundamental for a quantitative understanding of the structure and evolution of stars .
in fact , while the overall stellar structure is rather robust , changes of some of the rates by few percent can produce detectable discrepancies , when precise measurements are possible , _
e.g. _ , in the case of the solar photon and neutrino luminosity , and mechanical eigenfrequencies @xcite . in this section ,
we review the basic physics underlying thermonuclear reactions and the main ingredients of their calculation @xcite .
let us consider a gas with @xmath4 particles of type 1 and @xmath5 particles of type 2 per cubic centimeter and relative velocity @xmath6 ; the reaction rates @xmath7 ( the number of reactions per unit volume and unit time ) is given by @xmath8 where @xmath9 is the nuclear cross section of the reaction .
the reaction rate per particle pair is defined as @xmath10 where the particles distribution function @xmath11 is a local function of the temperature .
therefore , the reaction rate per particle pair @xmath12 is determined by the specific cross section and by the velocity distribution function of the incoming particles . in general , cross sections
do not have very strong dependence on the energy , when no energy barrier is present and away from resonances .
therefore , most of the contribution to @xmath12 comes from particles with energy of the order of @xmath13 , and the dependence on the specific form of @xmath11 is weak .
we shall see that the situation is very different in the presence of the coulomb barrier .
most of the nuclear reactions that power the stars are between charged particles .
these particle must penetrate a coulomb barrier that is very large in units of @xmath13 .
the penetration probability is proportional to the gamow factor @xmath14}$ ] , where @xmath15 is the sommerfeld parameter , @xmath16 is the fine structure constant and @xmath17 is the reduced mass ; this factor can be also written as @xmath18}$ ] , defining the gamow energy @xmath19 .
this exponentially small probability makes the cross section grow extremely fast with the energy ; therefore , one usually defines the astrophysical @xmath20 factor , whose energy dependence is weaker @xmath21 this reaction mechanism has at least two main consequences for the study of stellar structure .
\(1 ) since the bulk of particles in the stellar plasma have thermal energies of the order of @xmath2 , which is far below the coulomb barrier , only a small number of particles in the high - energy tail of the distribution has a chance of reacting : this high - energy tail plays a crucial rle for the reaction rates , which becomes very sensitive to small changes of the distribution .
\(2 ) the cross sections of some of the reactions , _ e.g. _ , weak reactions with coulomb barrier such as @xmath22 , are extremely small at the low energies relevant to stellar physics ( of the order of tens of kev for reactions in the sun ) .
therefore , some of them have only been calculated theoretically , others have only been measured at higher energies and then extrapolated to the lower thermal energies of the stellar interiors .
both these two facts lead to considerable uncertainties in the reaction rates .
the second point has been extensively discussed , and it is common opinion that the present standard theory of stars , and in particular of the sun @xcite , has already taken properly into account the uncertainties in the cross sections . on the contrary , the first point ,
the possibility that small changes of the particle energy distribution could strongly affect reaction rates , has received little attention @xcite , or dismissed on the ground that the energy distribution in stars is allegedly well - know @xcite . in this paper
, we shall discuss mainly this point , the energy distribution in stellar plasmas and its implication for nuclear rates .
normal stellar matter , such as the one in the sun , is nondegenerate , _
i.e. _ , quantum effects are small ( in fact , they are small for electrons and completely negligible for ions ) , it is nonrelativistic , and it is in good thermodynamical equilibrium . on this ground ,
the particle velocity distribution is almost universally taken to be a maxwell - boltzmann distribution , without much questioning .
however , derivations of the ubiquitous maxwell - boltzmann distribution are based on several assumptions @xcite . in a kinetical approach , one assumes ( 1 ) that the collision time be much smaller than the mean time between collisions , ( 2 ) that the interaction be sufficiently local , ( 3 ) that the velocities of two particles at the same point are not correlated ( boltzmann s stosszahlansatz ) , and ( 4 ) that energy is locally conserved when using only the degrees of freedom of the colliding particles ( no significant amount of energy is transferred to collective variables and fields ) . in the equilibrium statistical mechanics approach
, one uses the assumption that the velocity probabilities of different particles are independent , corresponding to ( 3 ) , and that the total energy of the system could be expressed as the sum of a term quadratic in the momentum of the particle and independent of the other variables , and a term independent of momentum , but if ( 1 ) and ( 2 ) are not valid the resulting effective two - body interaction is not local and depends on the momentum and energy of the particles . finally , even when the one - particle distribution is maxwellian , additional assumptions about correlations between particles are necessary to deduce that the relative - velocity distribution , which is the relevant quantity for rate calculations , is also maxwellian . at least in one limit
the mb distribution can be rigorously derived : systems that are dilute in the appropriate variables , whose residual interaction is small compared to the one - body energies . in spite of the fact that the effects of the residual interaction can not be neglected , as a good first approximation the solar interior can be studied in this dilute limit ; therefore , it is reasonable to suppose that the velocity distribution in the sun is not too far from the maxwellian one .
this fact is consistent with the many successes of the standard treatment and suggests that we start discussing the standard results and improve them later . in the ordinary treatment ,
the single particle energy distribution for protons and other ions is taken as @xmath23 when this distribution and a cross section of the form of eq .
( [ sigmas ] ) are inserted in eq .
( [ sigmav ] ) , the resulting integrand goes to zero both at large energies , because of the exponentially small number of particles , see eq .
( [ mbdis ] ) , and also at small energies , because of the exponentially small probability of barrier penetration , see eq .
( [ sigmas ] ) .
in fact , the integrand has a maximum at the temperature - dependent energy @xmath24 which it is called the most effective energy @xcite , since most of the particles that react have energies close to @xmath25 ; we have defined the adimensional parameter @xmath26^{1/3}$ ] .
figure [ figgamow ] gives a pictorial demonstration of how the gamow peak originates .
the exponentially decreasing function ( solid curve labeled `` maxwell '' ) is the maxwellian factor @xmath3 multiplied times 400 to emphasize the tail of the distribution .
the rapidly growing function ( dotted curve labeled `` penetration factor '' ) is the factor @xmath27 with the choice of @xmath28 , which corresponds to a most effective energy @xmath29 ; it has been arbitrarily normalized such that it is equal to 400 at @xmath30 .
the product of the two functions yields the gamow peak ( solid curve labeled gamow ) , which has been normalized to one at its maximum ( @xmath31 ) .
it is important to notice that the area under the maxwellian curve for energies larger than about @xmath32 ( where one finds most of the contribution to the peak ) is less then 0.3% of the total area . in this framework ,
one performs the integral in eq .
( [ sigmav ] ) using the saddle - point asymptotic expansion around the maximum @xmath25 of the gamow peak and finds that the reaction rate per particle pair is , apart for small calculable corrections , @xmath33 table [ table1 ] reports the values of @xmath25 in units of @xmath13 ( @xmath34 ) for several reactions .
we notice that @xmath25 can be much larger than @xmath13 and , therefore , only a very small number of particles in the far tail of the distribution contributes to the rate , _
e.g. _ , for the @xmath35he + @xmath35he reaction the most effective energy is about @xmath36 : there are less than about 40 particles out of a million that have energies so large or larger .
it is not sufficient anymore to know that the maxwellian distribution is good approximation , we must be sure that there are no corrections to a very high accuracy .
as discussed in sec . [ secrate ] , several subtle points must be assumed in the derivation of the often - taken - for - granted maxwellian velocity distribution .
likewise , one often assumes that the solar core could be treated as an ideal ( debye ) plasma .
however , there are physical conditions and/or specific applications that needs higher accuracy for which becomes necessary to take into account modifications of the standard plasma theory . in this section
we discuss how the physics of non - ideal plasmas , and in particular of the solar interior , can result in equilibrium velocity distributions that deviate from the `` standard '' one .
in addition , we present specific physical frameworks where non - maxwellian distributions arise and microscopic mechanisms that allow a reliable estimation of the size of deviations . in literature ,
a plasma is characterized by the value of the plasma parameter @xmath37 @xmath38 where @xmath39 is of the order of the interparticle average distance ( @xmath40 is the average density ) .
the plasma parameter is a measure of the ratio of the mean ( coulomb ) potential energy and the mean kinetic ( thermal ) energy .
depending on the value of the plasma parameter , we can distinguish three regimes that are characterized by different effective interactions and require different theoretical approaches .
* @xmath41 .
the plasma is described by the debye - hckel mean - field theory as a dilute weakly interacting gas .
the screening debye length @xmath42 is much greater than the average interparticle distance @xmath43 , hence there is a large number of particles in the debye sphere ( @xmath44 ) .
collective degrees of freedom are present ( plasma waves ) , but they are weakly coupled to the individual degrees of freedom ( ions and electrons ) and , therefore , do not affect their distribution .
binary collisions through screened forces produce the standard velocity distribution . *
the mean coulomb energy potential is not much smaller of the thermal kinetic energy and the screening length @xmath46 .
it is not possible to clearly separate individual and collective degrees of freedom .
the presence of at least two different scales of energies of the same rough size produces deviations from the standard statistics which describe the system in terms of a single scale , @xmath2 . * @xmath47 .
this is a high - density / low - temperature plasma where the coulomb interaction and quantum effects start to dominate and determine the structure of the system . in the solar interior
the plasma parameter is @xmath48 ; therefore , the solar core is a weakly non - ideal plasma where the debye - hckel conditions are only approximately verified .
similar behaviors are expected in other astrophysical systems characterized by plasma parameters in the range @xmath49 ; examples are brown dwarfs , the jupiter core and stellar atmospheres .
studies of systems with this intermediate values of @xmath37 are the most difficult ; it is possible to use / combine several different approaches none of which , however , is completely justified .
the reaction time necessary to build up screening after a hard collision can be estimated from the inverse solar plasma frequency @xmath50 sec , and it is comparable to the collision time @xmath51 sec .
therefore , several collisions are likely necessary before the particle looses memory of the initial state and the scattering process can not be considered markovian .
in addition , screening starts to become dynamical : the time necessary to build up again the screening after hard collisions is not negligible any more .
for concreteness we consider three possible approaches to weakly - noninteracting plasmas that yield nonstandard distributions .
these approaches are not alternative , but they could perhaps be seen as different and partial descriptions of the same complicated physical problem . the main purpose of their presentation is to give concrete examples of how nonstandard statistics arises .
however , they are not by any mean the last word on the topic : we must still develop their full potentiality and alternative approaches must also be pursued . in the fokker - plank context it is possible to introduce corrections ( i ) to the lowest order friction coefficient @xmath52 , and ( ii ) to the lowest order diffusion coefficient @xmath53 @xcite .
similar corrections have already been shown to exist in hydrodynamic systems .
_ we assume that the system is not too far from the standard regime that leads to the mb distribution , so that an expansion starting from the usual formalism makes sense_. the fokker - planck equation , given in the landau form , is @xmath54 where @xmath55 is the distribution probability of particles with velocity @xmath6 at time @xmath56 and @xmath52 and @xmath53 are the dynamical friction and diffusion coefficients .
the stationary distributions are the asymptotic solutions of the above equation .
to lowest order @xmath57 and @xmath58 , where the constant @xmath59 has dimension of time ( @xmath60 is the friction constant ) and @xmath61 has dimension of a velocity ( @xmath62 for brownian motion ) . at equilibrium one obtains the well - known maxwellian distribution @xmath63 we can generalize the standard brownian kinetics considering the expressions of the quantities @xmath52 and @xmath53 to the next order in the velocity variable : @xmath64 and @xmath65 ; these higher derivative terms can be interpreted as signals of nonlocality in the fokker - planck equation .
if @xmath66 and @xmath67 we find the tsallis distribution @xmath68^{1/(1-q ) } \theta\left[1 + ( q-1)\frac{m v^2}{2 kt}\right]\ , , \ ] ] where @xmath69 , @xmath70 is the heaviside step - function , and @xmath71 .
when the characteristic parameter @xmath72 is smaller than 1 ( @xmath73 ) , this distribution has a upper cut - off : @xmath74 ( the tail is depleted ) .
the distribution correctly reduces to the exponential maxwell - boltzmann distribution in the limit @xmath75 ( @xmath76 ) . when the parameter @xmath72 is greater than 1 ( @xmath77 ) , there is no cut - off and the ( power - law ) decay is slower than exponential ( the tail is enhanced ) .
if @xmath78 and @xmath79 , we find a druyvenstein - like distribution : @xmath80 which has also the functional form suggested by clayton @xcite to parameterize a small deviation ( depletion ) from the maxwellian statistics .
each particle is affected by the total electric field distribution due to the other charges in the plasma .
if external fields and large scale internal fields due to collective modes are subtracted from the total electric field , the single particle see the remaining field as a relative small random component .
the density of these random electric microfields has been studied @xcite and it is often expressed in terms of the dimensionless parameter @xmath81 as @xmath82 , where @xmath83 ; the distribution of @xmath81 in plasmas depends on the value of @xmath37 @xcite .
these microfields have in general long time correlations , and can generate anomalous diffusion .
the total ( micro)field can be decomposed into three main components . + ( i ) a slow - varying ( relative to the collision time ) component due to collective plasma oscillations , which the particle sees as an almost constant external mean field @xmath84 over several collisions .
+ ( ii ) a fast random component due to particles within a few debye radii , whose effect can be described by an elastic diffusive cross section @xmath85 . when only this cross section is present , the distribution remains maxwellian even in presence of the slow mean field @xmath84 .
+ ( iii ) a short - range two - body strong coulomb effective interaction , that can be described by the ion sphere model @xcite . the strict enforcement of this model as implemented by ichimaru @xcite yields the elastic cross section @xmath86 , where @xmath43 is the interparticle distance , @xmath16 an adimensional parameter whose order of magnitude can be inferred from the parameter @xmath81 that characterizes the microfields , @xmath87 . since @xmath88 , for @xmath89 , one can estimate @xmath90 . in the present model , it is this component of the electric field that turns out to be mainly responsible of the correction factor @xmath91}$ ] . in this framework , the stationary solution of the kinetic equation valid for small deviations from the mb distribution can be shown to be the druyvenstein distribution : @xmath92 } \ , , \label{fmicro}\ ] ] where @xmath93 @xmath94 @xmath95 is the elastic energy - transfer coefficient between two particles @xmath43 and @xmath96 , and @xmath97 is proportional to the square of the ratio of the energy densities of the electric field and of the thermal motion . in the small correction limit , relevant to the solar interior , @xmath98 and
the @xmath99 parameter is : @xmath100 from eq .
( [ fmicro ] ) , we see that the presence of electric microfields implies a deviation from the maxwell - boltzmann distribution and that the entity of this deviation depends on the value of the plasma parameter .
as already mentioned , in the solar core @xmath48 , hence , the order of magnitude of the deviation parameter @xmath99 is about @xmath101 and in this intermediate region long - range and memory effects take place .
as already discussed , in the solar core collective effects have time scales comparable to the average time between collisions ( _ e.g. _ , compare the inverse plasma frequency with the average collision time @xmath102 ) and it is not possible a description that separates the collective and the individual degrees of freedom .
a rigorous microscopic approach would imply the resolution of the dynamical equations of motion with a hamiltonian that explicitly contains collective and individual degrees of freedom and their mutual interactions .
our scope is more modest : we want to give a concrete example where memory effects ( long - time correlations ) are important .
the authors of ref .
@xcite introduce the collective variables @xmath103 and describe these collective variables as harmonic oscillator variables weakly coupled to the individual degrees of freedom , which , therefore , act as a thermostat for the collective variables .
generalizing this approach , we write down a generalized langevin equation for the collective variables @xmath104 where @xmath105 is the memory friction kernel , which takes into account the long - time tail of the correlations in the solar interior , and @xmath106 is the stochastic fluctuating thermal force due to the interactions with the individual degrees of freedom . in weakly non - ideal plasmas , the longitudinal waves have the following dispersion law @xmath107 the two functions @xmath105 and @xmath106 are not independent but related by the second fluctuation - dissipation theorem @xmath108 in the limit of negligible memory effects , the memory kernel becomes a @xmath97-function , that is , @xmath109 , where @xmath110 is the effective friction coefficient , which can be written in terms of the landau ( @xmath111 ) and collisional ( @xmath112 ) damping @xcite .
the explicit form of the memory kernel depends on the specific system and , in principle , should be deduced from a microscopic calculation . at the moment , we have been able to verify that , in the framework of tsallis nonextensive statistics , the memory kernel can be written , for @xmath113 , as @xmath114 the above expression is consistent with the prescription of the nonextensive statistics and implies anomalous subdiffusion for @xmath115 and superdiffusion for @xmath116 .
a more complete analysis of this possibility is still under way ; it is also important to study the effects of such a memory kernel for the collective modes on the individual degrees of freedom . at the moment
, we can speculate that such correlations among the collective modes could lead to a long - time asymptotical behavior of the velocity - correlation of the ions of the kind @xmath117 .
if @xmath118 , _ i.e. _ , the correlation decays sufficiently fast , the diffusion process is no qualitative different from the delta - function case : @xmath119 . in this case
, one can show @xcite that if @xmath120 and @xmath121 , the standard distribution of velocity is still valid .
however , if @xmath120 and @xmath122 ( or very small ) , or if @xmath123 , the diffusion is anomalous @xmath124 ( @xmath125 , if @xmath126 ) .
indeed , tsallis @xcite shows that the generalized entropy quite naturally generate both anomalous diffusion ( @xmath124 ) and the non - maxwellian probability distribution for the velocities of eq .
( [ tsdis ] ) . in this section
we have presented theoretical arguments that the velocity distribution in the solar core should deviate from the standard one and , in particular , that it could follow tsallis distribution @xcite shown in eq .
( [ tsdis ] ) .
we have also estimated the deviation from the standard statistics from the known microfield distribution .
if we use the parameterization suggested by clayton @xcite ( druyvenstein distribution ) @xmath127 @xmath99 should be of the order of 0.01 .
since in the limit @xmath75 the tsallis distribution can be asymptotically described as a druyvenstein with @xmath128 , we have also estimated that @xmath72 should be a few percent different from 1 . in the next section
, we shall demonstrate how even such small deviations from the maxwell - boltzmann distribution can be very important for solar physics .
both from general considerations about the successes of the standard approach and from the estimates and calculations that we have shown , it should be clear that the deviations from the maxwellian distribution in the sun are small .
therefore , it is completely general to make an asymptotic expansion of the following kind : @xmath129 where we have disregarded terms with powers higher than @xmath130 in the exponent and all other power corrections outside the exponential , apart the ones ( not shown ) needed to correctly normalize the distribution .
this result becomes the more accurate the more @xmath97 is small .
note that the exponential can not be expanded .
this same kind of parameterization was already considered by clayton @xcite .
one can easily convince oneself that the distributions previously discussed can be put in this form in the limit of small deviations .
we have seen that the the tsallis distribution plays a special rle : it can be also approximated to first order in @xmath131 by clayton s form with @xmath132 and a renormalized temperature @xmath133 . in sec .
ii we performed a saddle point expansion of the integral over the velocity distribution with the introduction of the most effective energy , @xmath25 , and approximating the integrand around the maximum @xmath25 with a gaussian function ( gamow peak )
. the same kind of asymptotic expansion can be repeated for this modified distribution yielding the following analytical expression for the change of the rate : @xmath134 where @xmath25 is the most effective energy of eq .
( [ mosteffen ] ) @xmath135 which depends on the reaction @xmath136 , through the charges @xmath137 and weights @xmath138 of the ions , and on the relevant average temperature @xmath139 ( @xmath140 kev is the temperature at the center of the sun ) , and where @xmath141 is the expression in eq .
( [ asyrate0 ] ) .
we shall use the expression in eq .
( [ asyratedel ] ) to discuss the effects of modified statistics on the solar thermonuclear rates : it has a sufficient numerical accuracy for this situation and it is physically more transparent .
however , we perform the integral over the distribution numerically in physical situations where stronger deviations from the maxwellian distribution are expected , and when we want to check whether the asymptotic expansion is sufficiently accurate . the dramatic effect of small deviations from the maxwellian distribution on the rate can be appreciated by looking at fig .
[ figgamow ] .
the steep decreasing dashed curve shows the high - energy tail of the tsallis distribution @xmath142^{1/(1 - q)}$ ] with @xmath143 compared with the corresponding maxwellian distribution ( solid curve ) .
the tails of the two distributions shown in fig .
[ figgamow ] contain about 0.30% ( maxwell ) and 0.26% ( tsallis ) of their total area and have been multiplied times a factor 400 to make their small difference more visible .
the product of each of these two very similar distributions times the same penetration factor ( dotted curve ) yields two very different gamow peaks .
the peak corresponding to the tsallis distribution ( dashed curve ) is not only shifted at lower energy by more than one unit of @xmath2 compared to the maxwellian peak ( solid curve ) , but it is also much lower : _ a tiny change of the distribution tail ( @xmath144 less probability in the tail ) makes the probability in the peak smaller by a factor of about 0.3 ! _
because of their small interaction cross section , neutrinos that reach the earth have not interacted with the solar matter and carry direct information from the solar core .
one might hope that their energy spectrum could tell us something about the particle velocity distribution in the sun .
the neutrino energy spectrum for a given reaction depends on the total energy available in the center of mass to the final products , total energy that is the sum of the thermal energy of the incoming particles and of the energy released by the reaction , and on the velocity of the center of mass that produces doppler broadening . as a specific example
, we shall use the @xmath1 fusion reaction , since its neutrino spectrum is easily calculable from phase space and since it releases a relatively low excess energy leaving more space for thermal effects .
the reaction @xmath145 releases an excess energy of @xmath146 mev , which is shared among the kinetic energies of the outgoing particles .
note the the total energy released by the reaction is @xmath147 mev , since the positron annihilation gives additional @xmath148 mev ; this energy , however , does not contribute to the kinetic energy of the neutrino ( this should be contrasted to the more rare reaction @xmath149 ) .
since there are three particles in the final state , the neutrino energy in the center of mass has a known probability distribution . for simplicity ,
we report only the result that disregards the recoil energy ( a relative correction of the order of @xmath150 ) @xmath151 the neutrino energy goes from zero to to the maximal energy @xmath152 where @xmath153 is the relative kinetic energy of the incoming protons , the approximate equality disregards the recoil energy consistently with eq .
( [ spectrpp ] ) , and where @xmath81 is the fermi function @xmath154 - 1 } \nonumber \\ & = & \frac { 2\pi\alpha ( k+q+m_{e}-e_{\nu } ) \left[(k+q -e_{\nu})(k+q+2m_{e}-e_{\nu})\right]^{-1/2 } } { \exp{\left[2\pi\alpha ( k+q+m_{e}-e_{\nu } ) \left[(k+q -e_{\nu})(k+q+2m_{e}-e_{\nu})\right]^{-1/2 } \right ] -1 } } \ , , \end{aligned}\ ] ] which takes into account the coulomb repulsion of positron and deuteron in the final state .
the neutrino energy spectrum is obtained by convoluting this result with the distribution for @xmath153 : @xmath155 . in fig .
[ figppsp ] , we show the energy distribution of the neutrinos from the @xmath1 reaction for three cases : ( 1 ) @xmath156 ( dotted curve ) , ( 2 ) @xmath157 kev with a maxwellian distribution ( solid curve ) and ( 3 ) @xmath157 kev with a @xmath158 tsallis distribution ( dashed curve ) . note that relative large value of @xmath159 are necessary in order that the effect on the spectrum of having a different distribution be of the same order of magnitude of the standard thermal effect . in principle , precise measurements of the neutrino spectrum give information on the thermal distribution of the proton kinetic energies @xmath160 .
let us estimate the necessary accuracy .
the central temperature of the sun is about @xmath161 kev and the most effective energy @xmath25 ( maximum of the gamow peak ) is about five times as big for the @xmath1 reaction ( see @xmath34 in table [ table1 ] ) ; therefore , thermal effects give a relative change of the maximal energy about @xmath162 and , in general , modify the shape of the spectrum by a few percent .
the center of mass motion gives also contributions of order @xmath2 .
spectral measurements with accuracy better than 1% start to be sensitive to the particle thermal motion and can measure the solar internal temperature .
however , measurements that are two orders of magnitude more accurate are necessary to measure the temperature with a few percent error and/or discriminate between the maxwellian distribution and those slightly distorted distributions ( @xmath163 or less ) that we shall consider in the next paragraph .
such kind of experiments are not expected to be feasible in the near future .
the solar neutrino problem is one of the most interesting long standing puzzle of the modern physics .
the combined results from the solar neutrino experiments ( homestake , gallex , sage , kamiokande and superkamiokande ) can not be reconciled with the predictions of the standard solar models ( ssm ) . since ssms have been very successful in predicting the stellar structure and have given excellent descriptions of measurements as detailed and accurate as the helioseismic ones , the neutrino experiments have suggested that the minimal standard electroweak model should be extended and include small neutrino masses and lepton flavor nonconservation @xcite .
neutrino oscillation theory would have far - reaching consequences for both particle physics and cosmology : therefore , it is of great importance to ask whether the solar neutrino problem can be solved , or at least alleviated , in the framework of the conventional physics @xcite
. the problem can be more precisely appreciated with the help of fig .
[ figsolnusp ] and table [ table3 ] .
figure [ figsolnusp ] show the energy spectrum of the main neutrino fluxes coming out of the sun .
the same figure shows also the part of the spectrum that contributes to each of the three present neutrino measures .
the energy dependence of each single flux depends practically only on nuclear physics and it is independent of the ssm ( see the discussion about the @xmath1 spectrum in the sec .
[ secspectrum ] ) . however , the integrated fluxes , which are reported in the first column of table [ table3 ] , and , therefore , the relative weight of the different components of the spectrum , depend on the ssm .
one should compare these predictions with the experimental data ( last column of table [ table3 ] ) .
only one experiment , the one performed by kamiokande and superkamiokande , ( detection of the cherencov light emitted in water by electrons that are elastically scattered by solar neutrinos , for the latest updated results see ref .
@xcite ) measures neutrinos from a single solar reaction ( @xmath164b @xmath165 , the so - called boron neutrinos ) , since its threshold ( see fig . [ figsolnusp ] ) is too high to see any of the other lower energy neutrinos .
therefore , its determination of the boron flux is solar - model independent ; its present result corresponds to less than one half of the predicted flux .
this experiment has measured also the energy spectrum of the boron neutrinos confirming the expected shape apart for same discrepancy for energies greater than about 14 mev , discrepancy that could be interpreted as evidence for oscillations @xcite .
the other two kinds of ( radiochemical ) experiments @xcite have lower thresholds and , since they do not have energy information , measure a combination of several neutrino fluxes .
their interpretation in terms of individual fluxes partially depends on the relative weight of the fluxes in the solar model .
however , they also show lower signals than predicted . for a review on the snp see refs .
@xcite .
the ssm flux predictions have ranges of variability , whose magnitudes mostly depend on their importance in the energy production mechanism .
the @xmath1 flux is almost independent of the ssm ( a few percent uncertainty ) , the @xmath166be is fairly stable ( about 10% uncertainty ) , while the boron flux , which is produced in a marginal chain of the energy production , has a larger ( more than 30% ) uncertainty . the ranges of predictions of the ssm fluxes come mainly from the experimental / theoretical uncertainties in the input parameters ( nuclear cross sections , photon opacities and initial contents ) .
the have been numerous attempts of reconciling solar models and neutrinos experiments by allowing all theoretical inputs to vary within and also considerably outside their uncertainties .
others have assumed arbitrary changes of the internal structure of the models ( scaling of the temperature profile , fast element mixing , etc . ) .
this approach to the snp is called the `` astrophysical solution '' ( references could be found in ref .
@xcite ) . the most radical _ astrophysical solution _ consists in leaving the neutrino fluxes as free parameters ( no solar model ) with the only constraint that the produced energy should match the solar luminosity . nevertheless , there remains a discrepancy at about the 34 @xmath167 level ( the central predictions of the ssm are more than @xmath168 s from the experimental results ) @xcite . a different velocity distribution yields a different @xmath169 , as discussed in sec .
[ secmodirate ] , and , therefore , a different solar model and different neutrino fluxes . in principle , the same variation of @xmath169 that is obtained by a different velocity distribution can also be obtained by an appropriate change of the cross section , for instance by means of the astrophysical factor @xmath20 in eq .
( [ sigmas ] ) .
therefore , allowing a modified velocity distribution does not span a different set of neutrino fluxes than the arbitrary variation of all the astrophysical factors .
in particular , one can not hope to achieve a better agreement with the data than the one achieved by leaving the fluxes themselves as free parameters . in practice , however , arbitrary large variations of the astrophysical factors contrast with our theoretical understanding of the nuclear reactions and often with experimental measurements . allowing generous , but not arbitrary ,
ranges for the input parameters , implies a much more restricted values of the possible neutrino fluxes . on the contrary
, small changes of the velocity distribution can give huge changes of @xmath169 .
therefore , changing the velocity distribution has several motivations .
( 1 ) when one is able to calculated the velocity distribution from the microscopic physics , it gives a predictions for @xmath169 .
( 2 ) even when one can only estimate the size of the deviation , it gives a physical mechanism to justify large and correlated changes of @xmath169 .
( 3 ) in any case , a given uncertainty in form of the distribution yields calculable greater ranges of solar model predictions compared to the ones of the ssm , which assumes the distribution function known to high accuracy .
changing @xmath169 for the @xmath136th reaction will affect the whole solar model and , in general , all fluxes will change .
we use eq .
( [ asyratedel ] ) as a general parameterization of the rate change for small deviations from the standard distribution .
the effects of modifying the rates on the fluxes can be estimated by using the power - law dependences @xmath170 for the fluxes @xmath171 @xmath166be , @xmath164b , @xmath172n and @xmath173o , and using the solar luminosity constraint @xcite to determine the @xmath1 flux , @xmath174 , and keeping fixed the ratio @xmath175 .
the exponents @xmath176 ( see table [ table2 ] ) have been taken from ref .
@xcite , where it is also discussed why solar models depend only on the combination @xmath177 and why it is a good approximation to keep the ratio @xmath178 constant .
a direct microscopic calculation would determine @xmath97 that could be different for every reaction ( @xmath179 ) .
the energy distribution can be influenced by the specific properties of the ion ( charge and mass ) and by the different conditions of the environment in those parts of the sun where each of the reactions mostly takes place .
however , such a direct calculation is not simple and it does not exist for the solar interior .
therefore , for the purpose of estimating the effect of nonstandard distributions , we consider two simple models and consider @xmath97 ( s ) as free parameter(s ) .
the first model assumes the same deviation @xmath97 for all distributions , the second model assumes that only the @xmath180be and @xmath35he + @xmath181he relative energy distributions are nonstandard and introduces @xmath182 and @xmath183 to parameterize their deviations . in the first case ,
one finds by substituting eq .
( [ edg ] ) into eq .
( [ dflux ] ) that @xmath184 where @xmath185 are reported in table [ table2 ] .
this dependence of the fluxes on @xmath97 is in good agreement with clayton s numerical calculation @xcite . using the model of ref .
@xcite as reference model and the experimental results up to the end of 1997 ( see table [ table3 ] ) , we obtain the best fit for @xmath186 with a @xmath187 . in the second case
, we proceed similarly , but we use @xmath182 for the reaction @xmath188be and @xmath183 for the reaction @xmath189he + @xmath190he : the corresponding @xmath191 and @xmath192 are also reported in table [ table2 ] . as shown in table [ table3 ] the best fit is obtained for @xmath193 ( negative @xmath97 corresponds to an enhanced tail , @xmath116 in tsallis distribution ) and @xmath194 with a @xmath195 . as expected from the discussion in the first parts of this section , this result is a solution to the snp , in the sense of providing a model to fit the experimental results within one ( _ a few _ ) sigma .
however , we have shown that deviations from standard statistics corresponding to values of @xmath97 of about 1% can change the neutrino fluxes of factors comparable to those that constitute the snp .
perhaps , the actual values of the neutrino fluxes coming out of the sun could result from the interplay of several mechanisms that are disregarded in the standard picture @xcite .
moreover , it is clear that the uncertainties of the neutrino fluxes are considerably underestimated by not considering the possibility of non - extensive distributions .
helioseismology provides very detailed and precise information on the solar structure .
the extremely precise measurements of a tremendous number of frequencies give the possibility of extracting values of the sound speed even near the solar core where the energy is generated .
in addition , several properties of the convective envelope are accurately determined .
it is important to verify that nonstandard distributions do not contrast with such data .
this study has been done for the velocity distribution of proton @xcite : if @xmath97 is within with following limits @xmath196 there is no incompatibility between helioseismic data and nonstandard statistics .
as we have seen , even such small values of @xmath97 have very important consequences for the neutrino fluxes ; in addition , this limit does not automatically apply to the distribution of other ions .
the solar core , where its energy is produced , is a weakly nonideal plasma .
many approaches are possible to such systems .
the standard debye - hckel theory is very successful , but it is not sufficient to give an accurate description when one needs to calculate the energy ( velocity ) distribution function to a high accuracy .
we have tried several new approaches to this complex problem .
in particular , we have considered corrections to the fokker - plank equations , the known stochastic distribution of electric microfields , and memory effects arising from the interaction of individual and collective variables .
all our attempts indicate that the velocity distribution of ions in the plasma could be different from the maxwellian one , and should be well described by a tsallis distribution slightly nonextensive ( @xmath197 ) . however , even such small deviations of the energy distribution produce dramatic effects on those nuclear rates whose main contributions come from the high - energy tail of the distribution , as it is best exemplified by fig .
[ figgamow ] .
y. suzuki ( superkamiokande coll . ) , in : neutrino 98 , proceedings of the xviii international conference on neutrino physics and astrophysics , takayama , japan , 4 - 9 june 1998 , y. suzuki and y. totsuka eds . to be published in nucl .
suppl . ) .
k. lande , in : neutrino 98 , proceedings of the xviii international conference on neutrino physics and astrophysics , takayama , japan , 4 - 9 june 1998 , y. suzuki and y. totsuka eds . to be published in nucl .
b ( proc . suppl . ) .
t. kirsten , in : neutrino 98 , proceedings of the xviii international conference on neutrino physics and astrophysics , takayama , japan , 4 - 9 june 1998 , y. suzuki and y. totsuka eds . to be published in nucl .
b ( proc . suppl . ) .
v. n. gavrin ( sage coll . ) , in : neutrino 98 , proceedings of the xviii international conference on neutrino physics and astrophysics , takayama , japan , 4 - 9 june 1998 , y. suzuki and y. totsuka eds . to be published in nucl .
b ( proc . suppl . ) . v.
gavrin _ et al . _
( sage collaboration ) , in _ neutrino 96 , proceedings of the 17th international conference on neutrino physics and astrophysics , helsinki , finland , 1319 june 1996 _ , edited by k. huitu , k/ enqvist , and j. maalampi ( world scientific , singapore , in press ) .
rcdd & & + @xmath198 : & @xmath199 & 4.8 & 23 .
+ @xmath200 : & @xmath201 & 13.8 & 190 .
+ @xmath202 : & @xmath203 & 16.8 & 281 .
+ @xmath204 : & @xmath205 & 17.4 & 303 .
+ @xmath206 : & @xmath207 & 20.2 & 407 . + .
the first four rows show @xmath208 , the logarithmic partial derivative of neutrino fluxes with respect to the parameter shown at the left of the row .
these numbers are discussed in ref .
the last three rows show @xmath209 , @xmath210 and @xmath211 , the logarithmic partial derivative of the fluxes with respect to the parameters @xmath97 s ; as discussed in the text , they are linear combinations of the @xmath16 s weighted by the factors @xmath110 of table [ table1 ] .
[ table2 ] [ cols="<,^,^,^ " , ] | the physics of nuclear reactions in stellar plasma is reviewed with special emphasis on the importance of the velocity distribution of ions .
then the properties ( density and temperature ) of the weak - coupled solar plasma are analysed , showing that the ion velocities should deviate from the maxwellian distribution and could be better described by a weakly - nonexstensive ( @xmath0 ) tsallis distribution .
we discuss concrete physical frameworks for calculating this deviation : the introduction of higher - order corrections to the diffusion and friction coefficients in the fokker - plank equation , the influence of the electric - microfield stochastic distribution on the particle dynamics , a velocity correlation function with long - time memory arising from the coupling of the collective and individual degrees of freedom . finally , we study the effects of such deviations on stellar nuclear rates , on the solar neutrino fluxes , and on the @xmath1 neutrino energy spectrum , and analyse the consequences for the solar neutrino problem . |
ultra - relativistic electrons accelerated through magnetic fields generate electromagnetic radiation , called synchrotron radiation ( sr ) . emission from sr sources may range from far infrared ( fir ) to x - ray frequencies .
the use of proper physical quantities and figures of merit enable their adequate characterization and deep understanding .
the properties of a sr source are usually described using three quantities : the spectral flux , the maximum of the angular spectral flux and the brightness @xcite-@xcite . among these quantities
the brightness has a very important role .
the physical meaning of brightness can be best understood by considering the imaging of the source on any experimental sample .
typically , only a small fraction of the photons in the beam can be effectively focused .
the brightness is an appropriate figure of merit for estimating the photon flux density on the sample , and was originally defined with the help of traditional radiometry .
traditional radiometry is based on a geometrical optics approach .
the basic quantity in this discipline is the radiance , which measures a spectral photon flux per unit area per unit projection solid angle .
since the radiance can be evaluated at any point along a photon beamline , it is associated with specific locations within an optical system , including an image plane where , usually , an experimental sample is placed .
all other radiometric units can be derived from the radiance integrating over area or solid angle . integrating over the solid angle yields a spatial flux . integrating over the area yields the angular flux . integrating over both area and solid angle yields total flux @xcite . in this picture , the radiance is the photon flux density in phase space . in non - dissipative cases
where liouville theorem holds , the radiance is an invariant quantity down the photon beamline .
the main issue with this concept stems from the fact that traditional radiometry only provides a natural description of the properties of light from incoherent sources .
second generation sr sources are characterized by poor transverse coherence , and application of the concept of radiance to sr allows for phase space analysis of the x - ray radiation . for these sources
the brightness is nothing more than the radiance .
however , with the advent of third generation sr sources , electron beams began to have ultra - small emittances and quantities used in traditional radiometry needed to be generalized to sr sources of arbitrary state of transverse coherence .
the question then arises whether it is possible to find a definition of brightness that , irrespective of the state of coherence of the sr source , has all properties that one normally associates with it in the geometrical optics limit .
the basic question concerning the characterization of third generation sr sources is the definition of brightness in terms of electromagnetic fields , and their statistical properties ( i.e on the basis of classical relativistic electrodynamics and statistical optics ) .
we begin our discussion with the generalization of the brightness definition first proposed by kim @xcite , which is essentially the wigner distribution ( wd ) @xcite of the sr electric field . in @xcite ,
the characteristics of the odd harmonics field from undulator sources in resonance approximation was analyzed in detail using the wd formalism .
the wd was explicitly derived both in the diffraction limited regime and in the electron beam size- and divergence - dominated regime .
after kim s pioneering papers @xcite no further theoretical progress was made .
in particular , the considerations by kim @xcite on bending magnet and wiggler brightness in terms of the wd formalism were not further developed into a satisfactory theory . in literature , the brightness of a sr source is sometimes defined as the wigner distribution itself , i.e. as a phase space quasi - probability function @xcite . as such , for sr sources of arbitrary state of transverse coherence , it is not guaranteed to be positive .
moreover it is convenient to introduce a figure of merit which always gives back a single , positive number and can serve , at the same time , as a measure for the wd . in this article
we shall define the brightness for any synchrotron source as the maximum value of the wd .
note that the brightness defined in this way is always a positive quantity .
we can regard it as a self - evident generalization of kim s choice of the on - axis peak value of the wd as a figure of merit for the undulator case but , surprisingly , we failed to find it in literature . in the following of this section we present an introductory discussion of brightness of synchrotron radiation sources in general .
then , in section [ oldres ] we give a summary of results in literature , and an overview of novel findings . in section [ sec : due ] we discuss an analysis of undulator brightness , and of the approximations proposed by kim @xcite . in the limiting cases where the geometrical optics treatment can still be applied , but only the electron beam size or the divergence dominate on the diffraction size and angle we find a significant numerical disagreement between exact and approximated calculations . in section [ sec : tre ]
we come to the new results and extend the wd formalism to a satisfactory theory of bending magnet brightness .
as already discussed , we will focus on the description of a sr source in terms of wd
. we will be interested in the case of an ultra relativistic electron beam going through a certain magnetic system .
we will discuss of an undulator to illustrate our reasoning , but the considerations in this section , being fully general , apply to any other magnetic system ( wiggler , bending magnet ) as well .
sr theory is naturally developed in the space - frequency domain , as one is usually interested into radiation properties at a given position in space at a certain frequency . in this article
we define the relation between temporal and frequency domain via the following definition of fourier transform pairs : @xmath3 we call @xmath4 the observation distance along the optical axis of the system and @xmath5 fixes the transverse position of the observer .
the contribution of the @xmath6-th electron to the field depends on the transverse offset @xmath7 and deflection angles @xmath8 that the electron has at some reference point on the optical axis @xmath4 e.g. the center of the undulator , that we will take as @xmath9
. moreover , the arrival time @xmath10 at position @xmath9 has the effect of multiplying the electric field in space - frequency domain by a phase factor @xmath11 , @xmath12 being the frequency .
any fixed polarization component of the total field in the space frequency domain , which is a scalar quantity , can therefore be written as @xmath13 where @xmath8 , @xmath7 and @xmath10 are random variables , and @xmath14 is the number of electrons in the bunch .
note that the single - electron field @xmath15 in eq .
( [ total ] ) is a complex quantity , and can be written as @xmath16 , with @xmath17 and @xmath18 real numbers .
it follows that the sr field at a fixed frequency and position is a sum of many phasors , one for each electron , of the form @xmath19 .
elementary phasors composing the sum obey three important statistical properties , that are satisfied in all sr problems of interest .
first , random variables @xmath10 are statistically independent of each other , and of variables @xmath8 and @xmath7 .
this assumption follows from the properties of shot noise in a storage ring , which is a fundamental effect related with quantum fluctuations .
second , the amplitudes @xmath20 have obviously finite mean @xmath21 and finite second moment @xmath22 .
third , we assume that the electron bunch duration @xmath23 is large enough so that @xmath24 . under these non - restrictive assumptions
the phases @xmath25 can be regarded as uniformly distributed on the interval @xmath26 . moreover , with the help of the central limit theorem , it can be demonstrated that real and imaginary parts of @xmath27 are distributed in accordance to a gaussian law .
it follows that sr is gaussian random process .
moreover , since one deals with pulses of finite duration , the process is intrinsically non - stationary .
an important consequence of the fact that the sr random process can be considered gaussian is that higher - order correlation functions can be expressed in terms of the second order correlation function with the help of the moment theorem @xcite . as a result
, the knowledge of the second - order correlation function in the space - frequency domain is all we need to completely characterize the signal from a statistical viewpoint .
the following definition holds : @xmath28 where brackets @xmath29 indicate ensemble average over electron bunches . for any given function @xmath30
, the ensemble average is defined as @xmath31 where integrals in @xmath32 and @xmath33 span over all offsets and deflections , and @xmath34 indicates the probability density distribution in the joint random variables @xmath8 , @xmath7 , and @xmath10 . the already discussed independence of @xmath10 from @xmath7 and @xmath8 allows to write @xmath35 as @xmath36 here @xmath37 is the longitudinal bunch profile of the electron beam , while @xmath38 is the transverse phase space distribution . substituting eq .
( [ total ] ) in eq .
( [ gamma ] ) one has @xmath39 } \right > .\cr & & \label{gamma2}\end{aligned}\ ] ] expansion of eq .
( [ gamma2 ] ) gives @xmath40 } \right\rangle \cr & & + \sum_{m\ne
n } \left\langle \bar{e}\left(\vec{\eta}_m,\vec{l}_m , z,\vec{r } _ { { } 1 } , \omega_1\right ) \exp{(i \omega_1 t_m)}\right\rangle\left\langle\bar{e}^ * \left(\vec{\eta}_n,\vec{l}_n , z,\vec{r } _ { { } 2 } , \omega_2\right ) \exp{(- i \omega_2 t_n ) } \right\rangle.\cr & & \label{gamma3}\end{aligned}\ ] ] with the help of eq . ( [ ensembledef ] ) and eq .
( [ independence ] ) one sees that the ensemble average @xmath41 is essentially the fourier transform ) coincides with our definition of temporal fourier transform , eq .
( [ ftdef2 ] ) , but it will not change if one uses another convention for the fourier transform definition . ] of the longitudinal bunch profile function @xmath37 , that is @xmath42 using eq .
( [ ftlong ] ) , eq .
( [ gamma3 ] ) can be written as @xmath43 where it is interesting to note that @xmath44 because @xmath45 is a real function .
when the radiation wavelengths of interest are much shorter than the bunch length we can safely neglect the second term on the right hand side of eq .
( [ gamma4 ] ) since the form factor @xmath46 goes rapidly to zero for frequencies larger than the characteristic frequency associated with the bunch length : think for instance , at a millimeter long bunch compared with radiation in the angstrom wavelength range ) is dominant with respect to the first , because it scales with the number of particles _ squared _ : in this case , analysis of the second term leads to a treatment of coherent synchrotron radiation ( csr ) phenomena . in this paper
we will not be concerned with csr and we will neglect the second term in eq .
( [ gamma4 ] ) , assuming that the radiation wavelength of interest is shorter than the bunch length .
also note that @xmath47 depends on the _ difference _ between @xmath48 and @xmath49 , and the first term can not be neglected . ] .
therefore we write @xmath50 as one can see from eq .
( [ gamma5 ] ) each electron is correlated just with itself : cross - correlation terms between different electrons was included in the second term on the right hand side of eq .
( [ gamma4 ] ) , which has been dropped .
on the one hand , for an electron bunch with rms duration @xmath23 the characteristic scale of @xmath51 is given by @xmath52 .
on the other hand , in all our cases of interest @xmath15 has a much slower dependence on @xmath12 than @xmath52 . as a result
, @xmath15 does not vary appreciably on the characteristic scale of @xmath51 and @xmath53 for bending magnet sources and undulator sources , respectively . here
@xmath54 is the number of undulator periods . in all cases of practical interest @xmath55 . ] .
we can therefore simplify eq .
( [ gamma5 ] ) to @xmath56 where @xmath57 is known as cross - spectral density . before proceeding we
introduce , for future reference , the notion of spectral degree of coherence , @xmath58 , that can be presented as a function of @xmath59 and @xmath60 as @xmath61^{1/2 } } = \frac{g(\vec{r}_1 , \vec{r}_2)}{\left[\langle |\bar{e}(\vec{r}_1)|^2\rangle \langle |\bar{e}(\vec{r}_2)|^2\rangle \right]^{1/2 } } , \label{gsmall}\end{aligned}\ ] ] where , for notational simplicity , we did not indicate the dependence of the single particle fields on @xmath62 and @xmath12 .
the function @xmath58 is normalized to unity by definition , i.e. @xmath63 .
( [ breakgamma ] ) fully characterizes the system under study from a statistical viewpoint .
correlation in frequency and space are expressed by two separate factors . in particular ,
spatial correlation is expressed by the cross - spectral density function under non - restrictive assumptions on the sr bandwidth and on the bunch duration . among all possible sr sources
, there exists an important class called quasi - homogeneous .
quasi - homogeneity is nothing but the spatial analogue of quasi - stationarity . in general , quasi - homogeneous sources are particular class of schell s model sources @xcite .
schell s model sources are defined by the condition that there is some position down the beamline , which we will call @xmath9 without loss of generality , such that their cross - spectral density ( note that eq .
( [ coore ] ) is defined at any position @xmath4 down the beamline ) is of the form @xmath64 equivalently , one may simply define schell s model sources using the condition that the spectral degree of coherence depends on the positions across the source only through the difference @xmath65 .
quasi - homogeneous sources are schell s sources obeying the extra assumption that @xmath66 varies so slowly with position that it is approximately constant over distances across the source , which are of the order of the correlation length , which is the effective width of @xmath67 . because of this , for quasi - homogeneous sources we are allowed to make the approximation @xmath68 where now @xmath69 is proportional to the intensity distribution at the source . by definition ,
quasi - homogeneous synchrotron sources obey a very particular kind of random process that is spatially ergodic .
qualitatively , a given random process is spatially ergodic when all ensemble averages can be substituted by 2d space averages . remembering this property one can derive the expression for the spectral degree of coherence @xmath70 which is the 2d autocorrelation function of a particular realization of the total electric field @xmath71 calculated at the source position . from this
we come to the conclusion , to be used in the following of this section , that the 2d fourier transform of @xmath67 is always a positive quantity as function of variables @xmath72 , @xmath73 is given by @xmath74 . here
@xmath75 is the fourier transform of @xmath76 as a function of the same variables @xmath72 and @xmath73 . ] . for our purposes
it is preferable to express the cross - spectral density in symmetrized form .
we therefore introduce the new variables @xmath5 , given by @xmath77 together with the previously defined difference @xmath78
@xmath79 we then have the inverse relations @xmath80 and @xmath81 if we now change variables in eq .
( [ coore ] ) according to eq .
( [ rave ] ) and eq , ( [ dr ] ) , we find that the general expression for the cross - spectral density can be written as @xmath82 let us consider a certain phase space distribution for an electron beam with a given transverse phase space distribution @xmath83 , which is a function of offset @xmath84 and deflection @xmath85 . at the source position one can write @xmath86 aside for a normalization constant @xmath87 , the inverse fourier transform of the cross - spectral density with respect to @xmath88 and @xmath89 can be written as : @xmath90 this is the expression for the wigner distribution in terms of the cross - spectral density .
we regard it as a distribution function defined over the four dimensions @xmath91 and parameterized by @xmath4 .
let us now specialize the expression in eq .
( [ wig1 ] ) to the case when the source is quasi - homogeneous .
since in that case the cross - spectral density function @xmath92 can be factorized into the product of intensity @xmath93 and spectral degree of coherence @xmath67 , the wigner distribution function also factorizes as @xmath94 having recognized that @xmath95 is proportional to the fourier transform of the spectral degree of coherence @xmath67 , and can be identified with the angular distribution of radiation intensity . since the distribution @xmath96 is the product of two positive quantities , never assumes negative values , and it can always be interpreted as a phase space distribution .
this analysis shows that quasi - homogeneous sources can always be characterized in terms of geometrical optics .
it also shows that in this particular case the coordinates in phase space , @xmath5 and @xmath97 , are separable . in the case of non quasi - homogeneous sources
one can still define @xmath96 using eq .
( [ wig1 ] ) . it can be shown that @xmath96 always assumes real values .
however , the wigner distribution is not always a positive function . as a consequence
, it can not always be interpreted as a phase space distribution . yet ,
the integral over @xmath5 and @xmath97 can be shown to be positive , and therefore the maximum of the wigner distribution is also bound to be positive , so that we can take this value as a natural definition for the brightness of sr sources .
the basic properties of the wigner distribution include statements about its 2d projections . if we make use of eq .
( [ wig1 ] ) we obtain the following expression for the @xmath98 projection : @xmath99 changing the order of integration and using the fact that the integral of the exponential function essentially results in a dirac @xmath100-function according to @xmath101 we see that @xmath102 in the quasi - homogeneous limit , as we just discussed , @xmath96 can be interpreted as photon distribution in phase space . then , for consistency with this limit , one should require that integrating the wigner distribution function over the solid angle @xmath103 yields the photon spectral and spatial flux density : @xmath104 where @xmath105 is electron beam current , @xmath106 is charge of the electron taken without sign , @xmath107 is the speed of light in vacuum and @xmath108 is the reduced planck constant .
here we have used parseval theorem , and included an additional factor two on the right - hand side of eq .
( [ intw ] ) , indicating that we use positive frequencies only .
comparison of the requirement in eq .
( [ intw ] ) with the mathematical property in eq .
( [ partialint ] ) fixes univocally the normalization constant @xmath87 as @xmath109 note that @xmath87 depends on the units chosen ( in this case gaussian units ) and on our definition of fourier transformation ( eq .
( [ ftdef2 ] ) ) .
the wigner distribution @xmath96 is also univocally defined as @xmath110 in the quasi - homogeneous asymptotic case , the brightness is a conserved quantity for perfect optical systems , and can be interpreted as maximum density of photon flux in phase space follows liouville equation .
this is noticeably the case for free - space . ] .
when one needs to specify the wigner distribution or the cross - spectral density at any position down the beamline , one needs to calculate the field at any position down the beamline . in order to do so , we first calculate the field from a single relativistic electron moving along an arbitrary trajectory in the far zone , and then we solve the propagation problem in paraxial approximation .
this last step allows calculation of the field at any position by backward - propagation in free - space with the help of the paraxial green s function , that is the fresnel propagator .
suppose we are interested in the radiation generated by an electron and observed far away from it . in this case
it is possible to find a relatively simple expression for the electric field @xcite .
we indicate the electron velocity in units of @xmath107 with @xmath111 , the lorentz factor ( that will be considered fixed throughout this paper ) with @xmath112 , the electron trajectory in three dimensions with @xmath113 and the observation position with @xmath114 . finally , we introduce the unit vector @xmath115 pointing from the retarded position of the electron to the observer . in the far zone , by definition ,
the unit vector @xmath116 is nearly constant in time .
if the position of the observer is far away enough from the charge , one can make the expansion @xmath117 we then obtain the following approximate expression for the the radiation field in the frequency domain ) compared to textbooks is to be ascribed to the use of different units and definition of the fourier transform . ] : @xmath118 \int_{-\infty}^{\infty } dt~{\vec{n}\times\left[\vec{n}\times{\vec{\beta}(t)}\right]}\exp \left[i\omega\left(t-\frac{\vec{n}\cdot \vec{r}(t)}{c}\right)\right ] ~,\cr & & \label{revwied}\end{aligned}\ ] ] where @xmath119 is the negative electron charge . using the complex notation , in this and in the following sections we assume , in agreement with eq .
( [ ftdef2 ] ) , that the temporal dependence of fields with a certain frequency is of the form : @xmath120 with this choice for the temporal dependence we can describe a plane wave traveling along the positive @xmath4-axis with @xmath121 in the following we will always assume that the ultra - relativistic approximation is satisfied , which is the case for sr setups . as a consequence , the paraxial approximation applies too .
the paraxial approximation implies a slowly varying envelope of the field with respect to the wavelength .
it is therefore convenient to introduce the slowly varying envelope of the transverse field components as @xmath122 introducing angles @xmath123 and @xmath124 , the transverse components of the envelope of the field in eq .
( [ revwied ] ) in the far zone and in paraxial approximation can be written as @xmath125 } } \left[\left({v_x(z')\over{c } } -\theta_x\right){\vec{e_x } } + \left({v_y(z')\over{c}}-\theta_y\right){\vec{e_y}}\right ] ~,\cr & & \label{generalfin}\end{aligned}\ ] ] where the total phase @xmath126 is @xmath127 \cr & & + \frac{\omega}{2c}\left[z_0 ( \theta_x^2+\theta_y^2 ) - 2 \theta_x x(z ' ) - 2 \theta_y y(z ' ) + z(\theta_x^2+\theta_y^2)\right]~ .
\label{totph}\end{aligned}\ ] ] here @xmath128 and @xmath129 are the horizontal and the vertical components of the transverse velocity of the electron , @xmath130 and @xmath131 specify the transverse position of the electron as a function of the longitudinal position , @xmath132 and @xmath133 are unit vectors along the transverse coordinate axis . finally , @xmath134 is longitudinal coordinate along the trajectory .
the electron is moving with velocity @xmath135 , whose magnitude is constant and equal to @xmath136 .
( [ generalfin ] ) can be used to characterize the far field from an electron moving on any trajectory . when the single - electron fields inside the ensemble average brackets @xmath29 are specified at a certain position @xmath137 , the fields at any other position @xmath138 can be found by propagating forward or backward in free - space according to the paraxial law @xmath139}~. \cr & & \label{fieldpropback}\end{aligned}\ ] ] in particular
, one may decide to backpropagate the field even at positions well inside the magnetic structure under study . in this case , the field distribution is obviously virtual in nature , because it is not actually there , but it fully characterizes the radiation field from a single electron with given offset and deflection . within the paraxial approximation , single - electron fields are fully characterized when they are known on a transverse plane at one arbitrary position @xmath4 . because of this , all positions @xmath4 are actually equivalent .
as we will see there can be , however , a privileged position @xmath140 where the electric field assumes a particularly simple form : at this position , in many cases of practical interest including undulator and bending magnet radiation , the field wavefront from a single electron is simply plane axis . ] . without loss of generality one can set @xmath141 for simplicity and call this the source position .
then , the relation between the field from a single electron at the source @xmath142 and the field in the far zone , @xmath143 , follows once more from eq .
( [ fieldpropback ] ) : @xmath144 @xmath145 we assume that a plane wave traveling along the positive @xmath4-axis can be expressed as in eq .
( [ eoftrav ] ) .
then , the negative sign in the exponential factor @xmath146 in eq .
( [ vtilde ] ) determines the sign of the exponential in eq .
( [ fieldpropback ] ) and consequently the sign of the exponential that appears in the integrand in eq .
( [ farzonef ] ) , which is the solution of the propagation problem in the far zone . returning to the definition of wigner distribution
, we see that in order to be able to give a physical interpretation to the wd variables @xmath72 and @xmath73 as angles of plane - wave propagation modes we must choose the same ( negative ) sign in the exponential that appear in eq .
( [ wig1 ] ) and in the integrand in eq .
( [ farzonef ] ) . if we insert eq .
( [ ggg2 ] ) into eq .
( [ wig13 ] ) we will obtain the most general expression for the wigner distribution , accounting for a given phase space distribution of the electron beam .
there are practical situations when offset and deflection of an electron lead to the same offset and deflection of the radiation beam from that electron .
this is the case for undulator and bending magnet setups without focusing elements . in such situations ,
the expression for the wigner distribution can be greatly simplified .
the presence of an electron offset @xmath84 shifts the single - electron field source , while a deflection @xmath85 tilts the source .
therefore @xmath147 \label{tiltshift}\end{aligned}\ ] ] where we set @xmath148 .
note that , due to duality between source and far zone plane , an electron offset tilts the single - electron far - zone field , while a deflection shifts the beam angle offset according to : @xmath149 \label{tiltshiftfar}\end{aligned}\ ] ] where we set @xmath150 . from eq .
( [ tiltshift ] ) one obtains @xmath151 where @xmath152 .
a very useful addition theorem @xcite can then be obtained from direct calculations : @xmath153 with @xmath154 defined as the wigner distribution associated to @xmath155 .
this can be summarized by saying that that electron offset and deflection correspond to an offset in position and angle of the corresponding wigner distribution @xmath154 , and that the overall wigner distribution @xmath96 can be found by addition over single - electron contributions .
we will apply this knowledge to the special cases of undulator and bending magnet sources respectively in section [ sec : due ] and section [ sec : tre ] . before that , in the next section we will present a summary of previous results and an overview of novel findings .
it is helpful to start our investigations by examining the brightness for quasi - homogenous sources .
the simplest case study is the undulator source . here
we take advantage of the particular but important situation of perfect resonance , when the undulator field can be presented in terms of analytical functions .
moreover , we will consider the practical case of a gaussian electron beam , which yields further analytical simplifications . in the quasi - homogeneous case of beam size- and divergence - dominated regime , the undulator brightness at the fundamental harmonic is easily shown to be @xmath156 where @xmath157 is the total spectral photon flux , and @xmath158 are the rms geometrical electron beam emittance in the horizontal and vertical directions .
usually the electron beam waist is located in the middle of the undulator , where @xmath159 , with @xmath160 the electron beam rms sizes and @xmath161 the electron beam rms divergences at that position .
therefore one obtains the well - known expression @xmath162 this result has , of course , a very simple physical interpretation in terms of maximum density of photon spectral flux in phase space .
in fact , in the beam size- and divergence - dominated regime the photon beam can be modeled as a collection of rays with the same phase space of the electron beam , and also the photon beam has its waist in the middle of the undulator .
another way of stating the same concept is by saying that the radiation source is located in the middle of the undulator .
one word of caution should be spent commenting this last result . from beam dynamics considerations
it is known that , if focusing elements are absent , the beam size varies along the undulator like @xmath163 where @xmath164 is the distance from the waist , with @xmath165 is the undulator length .
the average beam size along the undulator length is then @xmath166 textbook @xcite regards eq .
( [ text ] ) as a clear evidence of the fact that our undulator source possesses a finite longitudinal dimension : the size of this extended source changes along the undulator , and this effect tends to increase the effective source size and reduce the undulator brightness , as one can see from the average beam size eq .
( [ text ] ) , which is larger than that at the waist . however , this is a misconception because , according to the electrodynamics of ultrarelativistic charged particles , the source size is not widened at all , which is demonstrated by the fact that eq .
( [ bu1 ] ) for the brightness only includes the electron beam size at the waist . as discussed before ,
our source , placed in the middle of the undulator , is obviously virtual in nature .
however , it is no mathematical abstraction .
synchrotron radiation is often used to measure the size of the electron beam .
for example , in the electron beam size - dominated regime , undulator radiation can be used to form an image of the cross - section of the electron beam .
we can take a single focusing mirror of focal length @xmath167 at distance @xmath137 from the source to form a @xmath168 image at the same distance @xmath137 downstream of the lens .
we can set the object plane at the undulator center .
if , for example , the resolution is limited by diffraction effects only , the rms size of the image will be equal to the rms of the electron beam size at the waist , and it is not affected by variations of the parabolically shaped beta function along the undulator .
this image is also a visualization of the virtual source .
we should remark that statistical optics is the only mean to deal , in general , with the stochastic nature of sr . only in those particular cases
when sr can be treated in terms of geometrical optics beamline scientists can take advantage of ray - tracing techniques .
one of these cases is described above for the electron beam size- and divergence - dominated regime . in all generality , in order to decide whether geometrical optics or wave optics is applicable , one should separately compare the electron beam sizes and divergences at the electron beam waist position with the radiation diffraction sizes and diffraction angles , which are quantities pertaining the single - electron radiation .
when at least the electron beam size or divergence dominates , one can use a geometrical optics approach .
as we have seen , quasi - homogeneity is a necessary and sufficient condition for geometrical optics to be used in the representation of any sr source .
a source is quasi - homogeneous if and only if it is possible to factorize the cross - spectral density @xmath92 in the product of two factors separately depending on @xmath5 and @xmath78 .
for example , in the particular case when the electron beam size dominates over diffraction , the cross - spectral density @xmath92 ( and hence also the wigner distribution ) admits factorization , and the source can still be described with the help of geometrical optics even if the divergence is dominated by diffraction effects : in that case , since the source is quasi - homogeneous , the fourier transform of the spectral degree of coherence @xmath67 yields a diffraction limited intensity distribution in the far zone .
according to our definition , when we can treat sr in terms of geometrical optics , the brightness is always the maximum of the phase space density of the photon beam . in the beam
size - dominated or beam divergence - dominated limits , the undulator brightness can be determined analytically yielding the following cases ( see section [ sec : due ] for details ) : * beam divergence - dominated regime , for @xmath169 , @xmath170 . in this case
+ @xmath171 * beam size - dominated regime , for @xmath172 , @xmath173 . in this case
+ @xmath174 having dealt with quasi - homogeneous sources we now turn our attention to diffraction limited undulator sources . in the case of a zero - emittance electron beam , the undulator brightness can also be determined analytically yielding @xmath175 in literature it is often noted that the @xmath176 factor is the volume of a diffraction limited beam in the photon phase space .
the numerical factor four is dictated by the normalization condition eq .
( [ intw ] ) and by the axial symmetry properties of undulator radiation at the fundamental harmonic .
this point is discussed more in detail in @xcite . the following expression ( originally proposed by kim in @xcite )
is the usual estimate of the undulator brightness : @xmath177^{1/2 } } ~. \label{bundug}\end{aligned}\ ] ] eq .
( [ bundug ] ) can be obtained by approximating the radiation from a single electron by a gaussian laser mode with rms divergence and source size @xmath178 and @xmath179 .
then , one can use the addition theorem to obtain the brightness for a beam of electrons .
the integral in eq .
( [ wcorr ] ) , in this case , is a convolution of two gaussian functions . for large electron beam emittances
, this expression is in agreement with the geometrical optics limit pertaining to the beam size- and divergence - dominated regime , eq .
( [ bu1 ] ) .
once the electron sizes and divergences are fixed , in eq .
( [ bundug ] ) there are still two independent parameters , @xmath179 and @xmath178 . in @xcite ,
these two parameters are chosen in such a way that the diffraction limit in eq .
( [ bfil ] ) is satisfied .
this leaves only one degree of freedom to be fixed . in other words , according to eq .
( [ bundug ] ) , when the product @xmath180 is fixed , there is only one adjustable parameter to be fitted to the exact result for the brightness and @xmath181 . in this case
, the geometrical optics limit is reached for @xmath182 or @xmath183 . in section [ tgs ]
we will show that for both these asymptotes eq . ( [ bundug ] ) is in disagreement with the exact result .
] in the beam divergence - dominated regime , eq . ( [ bqh1 ] ) , and in the beam size - dominated regime , eq .
( [ bqh2 ] ) . in @xcite
the following definition fixes both the degrees of freedom available in eq .
( [ bundug ] ) : @xmath184 such choice has also been adopted in some articles @xcite and books @xcite . the approximate eq .
( [ bundug ] ) , together with the definitions in ( [ sigrrp0 ] ) , gives a qualitative agreement with exact results in all geometrical optics limiting cases , but detailed quantitative agreement is , in some case , rather poor .
in particular , it should be remarked that the approximate expression in eq .
( [ bundug ] ) overestimates the exact value of the brightness by eight times in the beam divergence - dominated regime , eq .
( [ bqh1 ] ) , and underestimates it by two times in the beam size - dominated regime , eq .
( [ bqh2 ] ) .
however , it should be noted that the choice of @xmath179 and @xmath178 in ( [ sigrrp0 ] ) is not the optimal choice ) , see e.g. @xcite .
the point here is that the exact expressions eq .
( [ bqh1 ] ) , eq .
( [ bqh2 ] ) are novel results that we could not find in literature .
because of this , only now we can optimize the last degree of freedom in eq .
( [ bundug ] ) by comparing the approximated expression with the exact results . ] for the approximation in eq .
( [ bundug ] ) .
in fact , the previous remark indicates that larger values of @xmath179 and smaller values of @xmath178 should be chosen .
more quantitatively , choosing @xmath185 we can obtain perfect numerical agreement with the exact result for the beam size - dominated regime eq .
( [ bqh2 ] ) and an overestimation of a factor four in the beam divergence - dominated regime , eq .
( [ bqh1 ] ) .
the choice of parameters in ( [ sigrrp ] ) was first introduced in @xcite and is used , nowadays , in most calculations exploiting the approximate eq .
( [ bundug ] ) @xcite . up to this point our analysis followed , in its general lines , the one given by kim in @xcite .
as the next step we turn to consider the brightness from bending magnets , which was not described , up to now , in a satisfactory way within the formalism of the wigner distribution .
this problem is still under discussion from the early days of sr theory @xcite , and it is reflected in the fact that many textbooks devoted to sr theory , like e.g. @xcite , do not discuss the brightness from bending magnets .
other books like @xcite discuss only the geometrical limit , and @xcite presents a qualitative treatment . only @xcite try to give a complete analysis of the problem .
part of the difficulty of applying the concept of brightness to the bending magnet case can be traced back to attempts using a mixture of intuitive geometrical optics and wave optics considerations , instead of exact results as was done in the undulator case .
a typical example of such intuitive description can be found in @xcite : `` to calculate the bending magnet brightness we need to consider the effective phase space which the photon flux is being emitted taking into account of both the finite electron and photon beam sizes and divergences .
first , there is no need to consider any horizontal angle effects as the light is emitted smoothly over the full horizontal @xmath186 radians '' . a quantitative definition of bending magnet brightness according to @xcite is then given by @xmath187 where @xmath188 is the photon flux per unit horizontal angle ( on the bending plane ) . the effective horizontal and vertical source size and effective vertical divergence
are then calculated to be @xmath189 where the vertical opening angle @xmath178 may be determined from the equality @xmath190 , @xmath191 being the on - axis photon flux density per unit solid angle @xcite . since ,
usually , only the horizontal polarization component of sr radiation from a bending magnet is important , here we discuss sr beam brightness only for the @xmath192 electric field component . using this method , considering the horizontally - polarized component of the field
one finds that @xmath193 at the critical wavelength @xmath194 , where @xmath195 is the radius of curvature of the electron orbit in the bending magnet times longer ) has been chosen so that this is a convenient combination to be used in section [ sec : tre ] for dimensional analysis of sr from a bending magnet in the space frequency domain . ] .
similarly to the undulator case , the angular divergence and source size for the radiation emitted by a single electron is chosen to satisfy @xmath196 .
it is instructive to examine this expression in diffraction limited case . from eq .
( [ btext ] ) we then find @xmath197 if the horizontal opening angle of the beamline , that is the half - angle subtended by the aperture in horizontal direction @xmath198 , is larger than @xmath178 , then the flux increases proportionally to @xmath198 .
however , in the limit for @xmath199 the brightness becomes independent of the beamline opening angle .
it is often noted that the factor @xmath176 is the minimal phase space volume of a diffraction limited beam .
therefore one can summarize by saying that sr emitted from a bending magnet within a solid angle of order @xmath200 occupies the minimal phase space volume @xmath176 .
the physical interpretation of the effect mentioned above can be found , among other references , in the review paper @xcite , where the brightness of the radiation from a bending magnet is described in the language of geometrical optics .
this language is intrinsically inadequate to describe the focusing of diffraction - limited radiation from a bending magnet .
in fact , this physical phenomenon fully pertains wave optics .
it is however possible to use geometrical optics reasoning and obtain an intuitive understanding of the situation . later on in this work we will show that such intuitive understanding is in qualitative agreement with an analysis fully based on wave optics . a complication in determining the brightness is the correlation between the longitudinal position of the source along the bending magnet and the ( horizontal ) angle of observation in the far zone .
an important parameter that is required to calculate the brightness is the source area .
in general , citing @xcite almost literally : `` there are three contributions to the horizontal source size and these are ( 1 ) the intrinsic size of the electron beam itself ( in our notation @xmath201 ) ; ( 2 ) the projected ( observed ) size due to the large horizontal sweep angle @xmath198 , leading to an extended source ( equal to the sagitta @xmath202 ) ; and ( 3 ) the diffraction limited source size . we can estimate the diffraction limited source size as @xmath203 '' .
the effective source size is then found by adding the three sizes in quadrature , thus @xmath204^{1/2 } ~. \label{totsize}\end{aligned}\ ] ] finally , it is worth emphasizing that if the horizontal opening angles @xmath198 is larger than the diffraction angle @xmath178 , the projected source size increases more rapidly @xmath205 than the flux @xmath206 @xcite . in literature
one can often find that the brightness is strictly linked to transverse coherence properties of the radiation and that the coherent flux @xmath207 can be defined as @xmath208 note that the common understanding that the coherence flux available after spatial filtering can always be written as eq .
( [ fcoh ] ) is erroneous .
for example , as we discussed above , the brightness of the radiation from a bending magnet in the diffraction limited case is constant when @xmath198 is larger than the diffraction angle @xmath178 .
however , in this situation the coherent flux increases proportionally to the horizontal opening angle @xmath198 . in fact , the entire photon flux collected from a diffraction limited source is fully transversely coherent . a young s double pinhole interferometer
can be used for demonstrating this fact . in the case of a diffraction limited source , the interference pattern recorded by the interferometer
is always characterized by a @xmath209 fringe contrast .
the physical meaning of brightness can be best understood by considering the imaging of the source on an experimental sample .
radiation is usually concentrated by using a wide - aperture focusing optical system .
the brightness is a figure of merit that quantifies how well a sr beam can be focused . for this purpose
one considers an ideal optical system .
in fact , in general , the maximum photon flux density onto the image plane is altered by optical elements along the setup .
we may interpret the brightness @xmath210 as the theoretical maximum concentration of the sr photon flux on the image - receiving surface where , usually , the sample is placed .
the source size can be affected by the presence of a finite electron beam .
it is often noted that the coherent part of the total flux can be ultimately focused down to a spot - size of dimension @xmath176 .
such consideration thus leads , at least qualitatively , to the relation between brightness and coherent fraction of total flux given by eq .
( [ fcoh ] ) .
it is clear that even a fully coherent beam can be of ` bad quality ' in the sense that a only small fraction of photons in the beam can be focused to a spot - size of dimension @xmath176 .
this is the case when the radiation source is characterized by a complicated wavefront , which has an effect similar to optics abberation , when there is a departure of the far field wavefront from the ideal spherical form .
then , one can not reach an effective focusing on the sample even in the diffraction limited case .
it is well - known that , in general , there are two different characteristics of the radiation beam .
the first one is the degree of transverse coherence @xmath211 , which reflects the statistical properties of radiation fields @xcite .
the second one is the well - known @xmath212 factor , which is widely used in the laser community to quantify how well a deterministic laser beam can be focused @xcite .
the diffraction - limited photon beam from an undulator has an @xmath212 factor close to unity and eq .
( [ fcoh ] ) is correct . at variance
, the diffraction - limited photon beam from a bending magnet has an @xmath212 factor close to unity only within a solid angle of about @xmath213 .
the brightness is a useful figure of merit that incorporates , simultaneously , both the statistical properties and the wavefront qualities of the radiation pulse .
when the electron beam has zero emittance we are dealing with perfectly coherent wavefronts . intuitively , in this situation one would apply methods from wave optics in order to solve the image formation problem . at variance , in literature it is often discussed an estimation of the source size for the case of diffraction limited radiation from a bending magnet based on geometrical optics @xcite .
a situation where one deals with a similar problem is in the calculation of a laser beam focus through a lens , when severe aberrations are present .
although the laser beam is coherent , when diffraction effects are negligible compared to aberration effects , the beam focusing can be calculated with the help of geometrical optics considerations . in the geometrical optics limit
, the wave equation can be replaced with the eikonal equation , which should be solved for surfaces of equal phase .
once the function of equal phase is known , one can apply usual ray - tracing techniques remembering that rays are , at any point , normal to the surface with equal phase .
the similarity between the two situations is highlighted by the essential feature of a diffraction limited sr beam from a bending magnet : at beamline opening angles @xmath199 wavefront ` aberrations ' are present in the sense discussed above , departure of the far field wavefront from the ideal spherical form can be considered as a coma - like aberration . ] , and are severe , meaning that @xmath214 .
therefore , a geometrical optics approximation , leading in particular to ray - tracing techniques , can be applied to the analysis of the image formation problem . before proceeding
we should make a few remarks concerning the terminology used in relation to brightness treatments .
the theory of brightness for a bending magnet is much more difficult than that for an undulator .
one can now see a net distinction between the geometrical optics limit in the framework of statistical optics when one discusses about an incoherent ( i.e. quasi - homogeneous ) sr source and the geometrical optics limit in the framework of coherent fourier optics , when one discusses about highly ` aberrated ' beams radiated from a single electron in the bending magnet setup .
an example where terminology is not accurately used can be found in @xcite : `` let us now turn to a more rigorous derivation of the source brightness of the synchrotron radiation due to a single electron . ... according to eq .
( 31 ) , photons are emitted incoherently in the tangential direction at each point of the trajectory '' .
it should be clear that one can talk about ` incoherently emitted ' photons only in the framework of statistical optics , when one deals with sr as random process . in the case of a single electron case we are always dealing with
coherently emitted photons at each point of the trajectory .
actually , the discussion in @xcite must be understood as an application of the geometrical optics approximation to the bending magnet radiation from a single electron in the case of large open angle , i.e when wavefront ` aberrations ' effects are dominant compared to diffraction effects . up to this point
our analysis of bending magnet brightness has followed that given in reviews @xcite and books @xcite on sr theory .
it is now instructive to consider our approach , and examine its results in the diffraction limited case . with our definition of brightness
we do not need to worry about how to account for the effect of sr beam ` aberrations ' .
in fact , the essential feature of our method is that we derive the brightness of a diffraction limited sr beam from the expression for the maximum wigner distribution .
clearly , the brightness defined in this way automatically includes a factor that characterizes the possibility of focusing the sr beam .
let us indicate with @xmath215 the brightness given by the approximated expression eq .
( [ bmx ] ) .
mathematically , our calculation is based on eq .
( [ wig13 ] ) .
one obtains that the bending magnet brightness in the diffraction limited case is given by @xmath215 times a function which contains the only variable @xmath216 .
in particular , at @xmath217 the exact result for @xmath218 is @xmath219 the approximated result that can be obtained from eq .
( [ bmx ] ) is naturally different from the exact one , although the difference is not large .
however , the fact remains that , at variance with the approximated case for undulator brightness , the usual estimate for the brightness of radiation from a bending magnet does not coincide with the exact result in the limit for a zero beam emittance .
the best way to avoid this kind of difficulties is to use the wigner function formalism as in the undulator case . only in this way it is possible to give an expression for the brightness of radiation from a bending magnet that is logically consistent and a directly applicable formulation that can be used by sr beamline scientists .
the most serious objection to approximation in eq .
( [ btext ] ) is that this expression does not include the electron beam divergence in the horizontal direction .
in fact , this is in contrast with results obtained within the wigner function formalism . in this respect ,
let us examine eq .
( [ btext ] ) in the following limiting case of beam divergence - dominated regime : @xmath220 eq .
( [ btext ] ) simplifies to @xmath221 which can also be written as @xmath222 here @xmath215 is given , as before , by eq .
( [ bmx ] ) . the contrast with the result obtained by exploiting the wigner function formalism can be seen by a straightforward application of eq .
( [ wig13 ] ) , which yields , in the same notations ( at @xmath217 ) : @xmath223 the horizontal electron beam divergence is a problem parameter .
we deduced this result by applying a rigorous mathematical method , without any intuitive arguments .
an intuitive way of understanding this property is to recall that due to wavefront ` aberrations ' , one has an @xmath212 factor close to unity only when considering radiation from a single electron emitted around the axis within a solid angle of about @xmath213 .
suppose that the electron beam has a divergence characterized by @xmath181 and @xmath183 . in this case , from geometrical considerations
it is evident that only a photon flux of order of @xmath224 can , in principle , be focused down to area an area of order @xmath225 .
therefore , the maximum photon flux density in phase space is proportional to the ratio of the flux in ( [ maxfoc ] ) and an effective phase space volume of order @xmath225 .
we thus deduced the parametric dependence of the bending magnet brightness in the limiting case ( [ siglim ] ) by means of intuitive arguments , which are in agreement with the rigorous mathematical derivation of eq .
( [ btocomp ] ) found within the wigner function formalism .
one can see that on the one hand the intuitively reasonable idea that there is `` no need to consider any horizontal angle effects as the light is emitted smoothly over full horizontal angle of @xmath186 radians '' @xcite can lead to incorrect results . on the other hand , however , other intuitive arguments are in agreement with the wigner function formalism .
only the existence of a rigorous , exact mathematical method guarantees a well - defined physical meaning for the brightness of a bending magnet .
another argument that disqualifies eq .
( [ btext ] ) as an approximation for bending magnet brightness follows from another comparison with exact results . in the beam size and divergence - dominated regime when @xmath226 , @xmath227 , @xmath183 and at arbitrary beam divergence in horizontal direction eq .
( [ btext ] ) yields : @xmath228 this last expression can be written as @xmath229 let us now consider rigorous calculations with the help of the wigner function formalism . as we discussed above , the beam size- and divergence - dominated regime is the simplest geometrical optics asymptote for the undulator case . on the contrary , for a bending magnet , it is the most complicated case to be treated analytically . intuitively we certainly expect that in this asymptotic limit there should be a competition between effects related with partial coherence ( since @xmath226 ) and what we called ` aberration ' effects ( since @xmath181 ) .
detailed mathematical analysis confirms such expectation .
it can be seen that in this case , depending on the specific ratio between horizontal beam size and divergence , the brightness is described using functions with completely different parametric dependence . in particular , eq .
( [ otherbug2 ] ) turns out to be parametrically inconsistent when the condition @xmath230 is satisfied .
it is interesting to see that intuitive and rigorous approaches coincide , at least in the limit when @xmath231 in fact , in this quasi - homogeneous case of beam size - dominated regime , which is typical for sr facilities in the x - ray wavelength range , from our definition of brightness we find @xmath232 with the help of simple algebra one can show that the brightness approximated by eq .
( [ btext ] ) actually coincides with eq .
( [ okour ] ) . although we can always find the brightness by a rigorous mathematical method following the wigner function approach , it is sometimes possible to get exact results such as eq .
( [ okour ] ) without any calculation .
the reasoning leading to such result is based on the fact that in the beam size - dominated regime the brightness is the product of the maximum angular flux density from a single electron and the maximum electron density at the source position . in other words , as already discussed , in the particular case of a quasi - homogeneous source , the brightness can always be interpreted in terms of geometrical optics , and the coordinates in phase space are separable ( see eq .
( [ wig2 ] ) ) .
the brightness is therefore a product of two positive quantities , eq .
( [ okour ] ) . the expression for the photon angular flux density radiated by a single electron in a bending magnet
is well known and can be found in any textbook devoted to sr theory .
the arguments we have just given for the bending magnet case can be applied to the undulator case as well . assuming a beam size - dominated regime and remembering the definition of brightness in the quasi - homogeneous limit as the maximum of the phase space photon flux density , one gets immediately that eq .
( [ okour ] ) gives the exact result eq .
( [ bqh2 ] ) . with this
, we point out an interesting fact about the angular flux density @xmath233 for the undulator radiation . if one calculates the total flux @xmath157 integrating over the solid angle , one obtains the relation @xmath234 , where @xmath235 is the on axis angular flux density in the diffraction limited case .
we can generalize the method for finding an exact result without any calculations to the asymptotic case where only the beam divergence is dominating . from the above analysis it is not hard to see that the brightness in this limit is given by the product of the maximum angular flux density of the electron beam and the maximum photon flux density radiated from a single electron at the source position .
in other words , when conditions ( [ siglim ] ) hold , for the divergence - dominated case we obtain in analogy with eq .
( [ okour ] ) : @xmath236 where @xmath237 is the maximum photon flux density on the source in the diffraction limited case . with eq .
( [ ok2 ] ) we can calculate the bending magnet brightness in the beam divergence - dominated regime at any wavelength .
all we need is an explicit expression for @xmath238 .
in contrast with the function @xmath233 in eq .
( [ okour ] ) , which is well - known , @xmath238 function is much less known . in the undulator case , to the authors knowledge , the only paper dealing with this issue is @xcite . for the bending magnet case
, we failed to find , in literature , an expression for @xmath238 .
an explicit formula for @xmath238 is derived in section [ sec : tre ] of this article . the previous analysis of the electron beam size- or divergence - dominated regime suggests an interesting question , whether it is always possible , in the geometrical optics limit , to deduce exact results for brightness without any calculations .
the answer is negative .
we could predict eq .
( [ okour ] ) and eq .
( [ ok2 ] ) based on intuitive grounds , and careful mathematical analysis confirms the expectation both for undulator and bending magnet cases .
let us now consider the case when electron beam size and divergence dominate . following intuitive arguments , the brightness should be proportional to the maximum of the electron beam density in phase space . from the analysis done above it is clear that this expectation is confirmed in the undulator case ( see eq .
( [ bu1 ] ) ) .
however , as described in section [ sec : tre ] , for a bending magnet in the beam size- and divergence - dominated regime the brightness is never inversely proportional to the product @xmath239 .
the theory of brightness for a bending magnet is much more difficult than that for an undulator , and only our exact mathematical method automatically includes all features that describe the possibility of focusing the sr beam .
we now focus on particular realizations of sr sources of practical interest . in particular , in this section
we consider the theory of brightness for undulator sources .
traditionally , all textbooks devoted to sr theory start discussing the characteristics of radiation from dipole magnets . only after a detailed study of this case they deal with other advanced topics like undulators . in this way
, the discussion follows the historical development of the subject , where undulators are presented as a logical extension of dipole magnets . however , the theory of bending magnet radiation is much more difficult than that of undulators .
therefore , here we choose to discuss undulators first .
we take advantage of the particular but important situation of perfect resonance , when the polarization direction does not depend on the observation angle and simply reproduces the polarization direction of the undulator field . in the far zone ,
undulator radiation from a single electron exhibits a finite divergence and a spherical wavefront centered in the middle of the setup .
in contrast , bending magnet radiation is emitted over the entire horizontal angle of @xmath186 radians and , moreover , the state of polarization of the radiation depends on the observation angle and does not exhibit an ideal spherical wavefront in the far zone .
these difficulties are reflected in the fact that bending magnet brightness was never described , up to now , in a satisfactory way within the wigner distribution formalism . in particular , in the usually accepted approximations , the description of the bending magnet brightness turns out to be inconsistent even qualitatively . at variance , the analysis of undulator brightness given by kim was based from the very beginning on the wigner distribution formalism .
this is reflected in the fact that the approximations he uses for describing the brightness are parametrically consistent with all exact results .
however , here we report numerical disagreement between exact results and approximated results in three geometrical optics limits where the brightness is nothing more than the maximum of the radiance , which is the photon flux density in phase space .
we consider a planar undulator , so that the transverse velocity of an electron can be written as @xmath240 where @xmath241 with @xmath242 the undulator period and @xmath243 the undulator parameter @xmath244 @xmath245 being the electron mass and @xmath246 being the maximum of the magnetic field produced by the undulator on the @xmath4 axis .
we will assume , for simplicity , that the resonance condition with the fundamental harmonic is satisfied .
in this way , our treatment leads to an analytical description of undulator radiation at the source position , i.e. in the middle of the undulator , at @xmath9 . the resonance condition with the fundamental harmonic
is given by @xmath247 a well - known expression for the angular distribution of the first harmonic field in the far - zone ( see appendix a for a detailed derivation ) can be obtained from eq .
( [ generalfin ] ) .
such expression is axis - symmetric , and can therefore be presented as a function of a single observation angle @xmath248 , where @xmath249 @xmath72 and @xmath73 being angles measured from the undulator @xmath4-axis in the horizontal and in the vertical direction .
one obtains the following distribution for the slowly varying envelope of the electric field : @xmath250 \mathrm{sinc}\left[\frac{\omega l\theta^{2}}{4 c}\right ] ~,\label{undurad4bis}\end{aligned}\ ] ] where the field is polarized in the horizontal direction . here
@xmath251 is the undulator length and @xmath54 the number of undulator periods .
finally , @xmath252 is defined as @xmath253 @xmath254 being the n - th order bessel function of the first kind .
eq.([undurad4bis ] ) describes a field with spherical wavefront centered in the middle of the undulator .
( [ farzone ] ) can now be used to calculate the field distribution at the virtual source yielding @xmath255~ , \label{undurad5}\end{aligned}\ ] ] where @xmath256 indicates the sin integral function and @xmath257 is the distance from the @xmath4 axis on the virtual - source plane .
note that @xmath258 is axis - symmetric .
( [ undurad5 ] ) , that has been already presented in @xcite , describes a virtual field with a plane wavefront .
let us compare this virtual field with a laser - beam waist . in laser physics , the waist is located in the center of the optical cavity . in analogy with this , in our case
the virtual source is located in the center of the undulator .
both in laser physics and in our situation the waist has a plane wavefront and the transverse dimension of the waist is much longer than the wavelength .
note that the phase of the wavefront in eq .
( [ undurad5 ] ) is shifted of @xmath259 with respect to the spherical wavefront in the far zone .
such phase shift is analogous to the guoy phase shift in laser physics .
finally , in laser physics , the rayleigh range for a laser beam is presented in the form @xmath260 , @xmath261 being the radius of the beam at the location of the waist ( i.e. at that position along @xmath4 where the wavefront is flat ) .
this is defined , for example , by requiring that the intensity on the edge of an aperture of radius @xmath261 be one fourth of the intensity at the center of the radiation spot . in the undulator source case , the definition given above amounts to @xmath262 and @xmath263 . in the case of a laser beam
the rayleigh range is related to the resonator geometrical factor . in analogy with this , in the case of an undulator source the rayleigh range is related to the undulator geometrical factor .
the relative intensity at the virtual source is plotted in fig .
[ virundu ] .
( [ undurad4bis ] ) and eq .
( [ undurad5 ] ) can be generalized to the case of a particle with a given offset @xmath84 and deflection angle @xmath85 with respect to the longitudinal axis , assuming that the magnetic field in the undulator is independent of the transverse coordinate of the particle .
although this can be done using eq .
( [ generalfin ] ) directly , it is sometimes possible to save time by getting the answer with some trick .
for example , in the undulator case one take advantage of the following geometrical consideration @xcite , which are in agreement with rigorous mathematical derivation .
first , we consider the effect of an offset @xmath84 on the transverse plane , with respect to the longitudinal axis @xmath4 .
since the magnetic field experienced by the particle does not change , the far - zone field is simply shifted by a quantity @xmath84 .
( [ undurad4bis ] ) , can be immediately generalized by systematic substitution of the transverse coordinate of observation , @xmath264 with @xmath265 .
this means that @xmath266 must be substituted by @xmath267 , thus yielding @xmath268 \mathrm{sinc}\left[\frac{\omega l \left|\vec{\theta}-\left({\vec{l}}/{z_0}\right)\right|^2}{4 c}\right ] ~.\label{undurad4bisgg0}\end{aligned}\ ] ] let us now discuss the effect of a deflection angle @xmath85 .
since the magnetic field experienced by the electron is assumed to be independent of its transverse coordinate , the trajectory followed is still sinusoidal , but the effective undulator period is now given by @xmath269 .
this induces a relative red shift in the resonant wavelength @xmath270 . in practical cases of interest
we may estimate @xmath271 .
then , @xmath272 should be compared with the relative bandwidth of the resonance , that is @xmath273 , @xmath54 being the number of undulator periods .
for example , if @xmath274 , the red shift due to the deflection angle can be neglected in all situations of practical relevance . as a result
, the introduction of a deflection angle only amounts to a rigid rotation of the entire system . performing such rotation
we should account for the fact that the phase factor in eq .
( [ undurad4bisgg0 ] ) is indicative of a spherical wavefront propagating outwards from position @xmath9 and remains thus invariant under rotations .
the argument in the @xmath275 function in eq .
( [ undurad4bisgg0 ] ) , instead , is modified because the rotation maps the point @xmath276 into the point @xmath277 . as a result , after rotation , eq .
( [ undurad4bisgg0 ] ) transforms to @xmath278 \mathrm{sinc}\left[\frac{\omega l \left|\vec{\theta}-\left({\vec{l}}/{z_0}\right)-\vec{\eta}\right|^2}{4 c}\right ] \cr & & \label{undurad4bisgg00}\end{aligned}\ ] ] finally , in the far - zone case , we can always work in the limit for @xmath279 , that allows one to neglect the term @xmath280 in the argument of the @xmath275 function , as well as the quadratic term in @xmath281 in the phase .
( [ undurad4bisgg00 ] ) can be further simplified , giving the generalization of eq .
( [ undurad4bis ] ) in its final form : @xmath282 \mathrm{sinc}\left[\frac{\omega l \left|\vec{\theta}-\vec{\eta}\right|^2}{4 c}\right ] ~. \label{undurad4bisgg}\end{aligned}\ ] ] the expression for the field at virtual source , eq .
( [ undurad5 ] ) , should be modified accordingly .
namely , one has to plug eq .
( [ undurad4bisgg ] ) into eq .
( [ farzone ] ) , which gives @xmath283 \mathrm{sinc}\left[\frac{\omega l \left|\vec{\theta}-\vec{\eta}\right|^2}{4 c}\right ] \cr & & \label{undurad4bisggtr}\end{aligned}\ ] ] yielding @xmath284\left[\pi - 2\mathrm{si } \left(\frac{\omega \left|\vec{r}-\vec{l}\right|^2}{l c}\right)\right]~\cr & & \label{undurad5gg}\end{aligned}\ ] ] as final result .
the meaning of eq .
( [ undurad5gg ] ) is that offset and deflection of the single electron motion with respect to the longitudinal axis of the system result in a transverse shift and a tilting of the waist plane .
the combination @xmath285 in eq .
( [ undurad5gg ] ) describes the shift , while the phase factor represents the tilting of the waist plane . to sum up ,
the diffraction size of the undulator radiation beam is about @xmath286 .
this means that the radiation from an ultra - relativistic electron can be interpreted as generated from a virtual source , which produces a laser - like beam . in principle
, such virtual source can be positioned everywhere down the beam , but there is a particular position where it is similar , in many aspects , to the waist of a laser beam . in the case of an undulator
this location is the center of the insertion device , where virtual source exhibits a plane wavefront . for a particle moving on - axis ,
the field amplitude distribution at the virtual source is axially symmetric , eq . ( [ undurad5 ] ) .
when the particle offset is different from zero , the laser - like beam is shifted .
when the particle also has a deflection , the laser - like beam is tilted , but the wavefront remains plane .
then , since radiation from a given electron is correlated just with itself , it follows that radiation from an electron beam is an incoherent collection of laser - like beams with different offsets and deflections . for a filament electron beam of current @xmath105
, the angular spectral flux density in the direction @xmath287 can be written as @xmath288 where @xmath289 is the number of photons per unit time per unit solid angle per relative frequency bandwidth , and @xmath290 is the slowly varying envelope of the electric field produced by a single electron in a planar undulator at the resonance wavelength in the space - frequency domain , eq .
( [ undurad4bis ] ) .
( [ dfdom ] ) can be directly deduced from eq .
( [ intw ] ) considering that @xmath291 .
it can be shown from that the maximum value of @xmath292 as a function of @xmath72 and @xmath73 is reached on - axis , for @xmath293 .
the angular spectral flux on - axis is given by @xmath294 where @xmath295 is the fine structure constant .
the angle - integrated spectral flux @xmath296 is defined as @xmath297 if we substitute eq .
( [ dfdom ] ) in eq .
( [ angint ] ) we obtain @xmath298 from eq .
( [ fsub ] ) and eq .
( [ maxdfdom ] ) one derives the following useful relation between the on - axis angular spectral flux and the angle integrated spectral flux @xmath299 the results presented above for our undulator source analysis is far from trivial . as a matter of fact ,
our results are in contrast with literature .
let us consider as cardinal example the idea that the calculation of the undulator brightness needs to take into account depth - of - field effects ( i.e. the contribution to the apparent source size from different poles ) .
this idea can be found , for example , in @xcite and is a misconception originating from the analysis of the undulator source based on intuitive geometrical arguments regarded as self - evident .
for example , in the textbook @xcite one can find : `` the actual photon brightness is reduced from the diffraction limit due to betatron motion of the particles , transverse beam oscillation in the undulator , apparent source size on axis and under an oblique angle .
all of these effects tend to increase the source size and reduce brightness .
the particle beam cross section varies in general along the undulator .
we assume here for simplicity that the beam size varies symmetrically along the undulator with a waist in its center . from beam dynamics it
is then known that , for example , the horizontal beam size varies like @xmath300 , where @xmath301 is the beam size at the waist , @xmath302 the divergence of the beam at the waist and @xmath303 the distance from the waist .
the average beam size along the undulator length @xmath165 is then '' given by eq .
( [ text ] ) . " similarly , due to an oblique observation angle @xmath248 with respect to the @xmath304-plane or @xmath305 with respect to the @xmath306-plane we get a further additive contribution @xmath307 to the apparent beam size . finally , the apparent source size is widened by the transverse beam wiggle in the periodic undulator field .
this oscillation amplitude is @xmath308 . collecting all contributions and adding them in quadrature , the total effective beam - size parameters are given by @xmath309 where
the particle beam sizes can be expressed by the beam emittance and betatron function as @xmath310 , @xmath311 , and the diffraction limited beam parameters are @xmath312 and @xmath313 " .
we stress that our criticism is not only focused on the quantitative aspects the source size analysis in @xcite , but rather on a contradiction with fundamental facts from classical electrodynamics . in order to better see this , consider an optical setup performing a 1:1 imaging , where we set the object plane at the undulator center .
the intensity distribution in the image plane is given by a convolution of the intensity from a single electron , @xmath314 , with the electron density distribution in the object plane ( which we consider gaussian , as before ) : @xmath315~. \label{intensity}\end{aligned}\ ] ] here @xmath316 ^ 2 $ ] is the radiation intensity from a single electron on the source ( see eq .
( [ undurad5gg ] ) ) .
( [ intensity ] ) follows from statistical averaging over electron offsets and deflection angles .
in fact , tilt angles are small compared to unity and , therefore , the source size is insensitive to the angular beam distribution .
it is easy to see that the first two terms in eq .
( [ wied12 ] ) for the source size are related with the approximation of the convolution eq .
( [ intensity ] ) introduced by kim @xcite .
however , the last two terms in eq .
( [ wied12 ] ) , which describe a source size widening due to depth - of - field effects , depend on the angular beam distribution and observation angle , and should not be there .
they are rather the result of the misconception discussed above .
there is another objection that could be made to the analysis in @xcite , related with the third term in eq .
( [ wied12 ] ) . under the resonance approximation
it is logically inconsistent to present a term of order of @xmath317 in the expression for undulator source size .
in fact , on the one hand the radiation diffraction size for a single electron is order of @xmath318 . on the other hand
the electron wiggling amplitude is given by @xmath319 .
in particular , at the fundamental harmonic follows that @xmath320 where we used the resonance condition eq .
( [ rsfirsth ] ) .
this inequality holds independently of the value of @xmath243 , because under the resonance approximation @xmath321 .
thus , the electron wiggling amplitude is always much smaller than the radiation diffraction size .
it follows that under the resonance approximation the third term in eq .
( [ wied12 ] ) should be neglected .
note that the first - principle computer codes ( see e.g. @xcite ) have been used quite successful to model advanced sr sources and beamlines without specific analytical simplifications .
results may be obtained using numerical techniques alone , starting from the lienard - wiechert expression for the electromagnetic field ( in the case of @xcite in the space - frequency domain ) using only the ultrarelativistic approximation .
codes can also be used to treat the case of 1:1 imaging of an undulator source .
it is instructive to reconsider the problem of undulator source size prediction by means of numerical simulations alone , which play the same role of an experiment .
we might consider two cases at fixed electron beam size : * the usual case with matched beta - function @xmath322 , and additionally @xmath323 , @xmath324 , @xmath226 . * a case with a tenfold increase in emittance and a tenfold decrease in beta function i.e. @xmath325 and @xmath326
. of course one is free to choose other numerical cases as well .
one can check that results of simulations confirm our prediction that source image is insensitive to the electron beam divergence . in both cases , with graphical accuracy , we find that the distribution at the image plane is the same , in contrast with the prediction @xcite of source size widening .
we now turn to the main topic of this study , namely the analysis of the brightness of sr sources . in this section
we apply the considerations developed in section [ sec : wiggen ] to the case of an undulator source at resonance .
first we calculate the wigner distribution for a filament electron beam , that is a beam with zero emittance . as second step we take into account the more general case when the electron beam has a finite phase space distribution . according to the definition in eq .
( [ wig13 ] ) , the wigner distribution in the case of an electron beam with zero emittance is given by @xmath327 where @xmath328 is the diffraction - limited cross - spectral density and @xmath329 is defined by eq .
( [ undurad5 ] ) .
the peak value of @xmath154 as a function of @xmath5 and @xmath97 is reached on - axis for @xmath330 and @xmath331 .
accordingly , we can compute the undulator brightness . for our diffraction - limited regime at resonance
this can be done analytically @xcite @xmath332 where @xmath157 is the angle - integrated spectral flux defined by eq .
( [ fsub ] ) .
here we have only used the axial symmetry of the electric field radiated by a single electron in an undulator at resonance , eq .
( [ undurad5 ] ) . in
particular this symmetry yields the relation @xmath333 .
it is reasonable to expect that eq .
( [ brightund1 ] ) is valid for any radiation beam with axial symmetric field distribution .
one can easily show that this is indeed the case .
one typical example when this fact is verified , is for a gaussian beam .
the wigner distribution @xmath334 for an electron beam with finite emittance can be presented as a convolution product between the electron phase space distribution @xmath335 and the wigner distribution for a filament beam @xmath336 according to eq .
( [ wcorr ] ) @xcite .
note that , as remarked before , eq .
( [ wcorr ] ) has no full generality and can be used only in the case when focusing elements are excluded from consideration . in the following
we will make consistent use of dimensional analysis , which allows one to classify the grouping of dimensional variables in a way that is most suitable for subsequent study .
normalized units in the undulator case will be defined as @xmath337 we assume that the motion of electrons in the horizontal and vertical directions are completely uncoupled . additionally we assume a gaussian distribution of the electron beam in phase space .
these two assumptions are practically realized , with good accuracy , in storage rings . for simplicity
, we also assume that the minimal values of the beta - functions in horizontal and vertical directions are located at the middle of the undulator , at @xmath338 .
then , at that position , the transverse phase space distribution can be expressed as @xmath339 with @xmath340 here @xmath341 parameters @xmath342 will be called beam diffraction parameters , are analogous to fresnel numbers and correspond to the normalized square of the electron beam sizes , whereas @xmath343 represent the normalized square of the electron beam divergences .
we begin by writing the expression for the cross - spectral density at the virtual source : @xmath344 where the field is defined by eq .
( [ undurad5 ] ) .
it should be noted that the independent variable in eq .
( [ undurad5 ] ) is @xmath345 , and corresponds to @xmath346 in dimensionless units .
thus , the characteristic transverse range of the field at the source in dimensionless units is of order of unity .
one sees that the cross - spectral density is the product of two separate factors .
the first is the fourier transform of the distribution of the electrons angular divergence .
the second is the convolution of the transverse electron beam distribution with the four - dimensional function @xmath347 .
in fact , after the change of variables @xmath348 we have @xmath349 \exp \left[-\frac{(\delta \hat{y})^2 d_y}{2}\right]\cr & & \times \int_{-\infty}^{\infty } d \phi_x \int_{-\infty}^{\infty } d
\phi_y \exp\left[-\frac{\left(\phi_x-\hat{x}\right)^2}{2 n_x}\right ] \exp\left[-\frac{\left(\phi_y-\hat{y}\right)^2}{2 n_y}\right ] \cr & & \times \widetilde{e}\left({\phi_{x}}+\frac{\delta \hat{x}}{2},{\phi_{y}}+\frac{\delta \hat{y}}{2 } \right ) \widetilde{e}^*\left({\phi_{x}}-\frac{\delta \hat{x}}{2 } , { \phi_{y}}-\frac{\delta \hat{y}}{2 } \right)~.\label{gnor3}\end{aligned}\ ] ] it is instructive to examine this expression in the geometrical optics asymptotes . let us start with beam size- and divergence - dominated regime . in eq .
( [ gnor3 ] ) the range of the variable @xmath350 is effectively limited up to values @xmath351 .
in fact @xmath350 enters the expression for @xmath290 .
it follows that at values larger than unity the integrand in eq .
( [ gnor3 ] ) is suppressed .
moreover , in the beam size- and divergence - dominated regime one has @xmath352 and @xmath353 , so that we can neglect @xmath350 in the exponential functions in @xmath342 , while from the exponential in @xmath343 follows that @xmath354 and @xmath355 can be neglected in the field @xmath290 . as a result
we obtain @xmath356 or , in dimensional units , @xmath357 one can obtain the wigner distribution @xmath96 from the above expression for @xmath92 by means of eq .
( [ wig13 ] ) @xmath358 the peak value of the wigner distribution is reached on - axis and is given by @xmath359 where @xmath157 is the spectral flux @xmath360 radiated by an electron beam with current @xmath105 in the undulator .
let us now consider the beam divergence - dominated regime , that is when @xmath361 . from an analysis of the exponential functions in @xmath343 in eq .
( [ gnor3 ] ) follows that @xmath354 , @xmath355 can be neglected in the expression for the field @xmath290 .
then , since @xmath362 , it follows that the distributions @xmath363 $ ] and @xmath364 $ ] are respectively sharply peaked about @xmath365 and @xmath366 .
hence , we can approximate @xmath367 finally , from eq .
( [ wig13 ] ) follows an expression for @xmath96 in dimensional units : @xmath368 where @xmath329 is the radiation field in eq .
( [ undurad5 ] ) .
the peak value of wigner distribution is reached on - axis and given by @xmath369 where @xmath370 is the maximum photon flux density on the source in the diffraction limited case .
a straightforward integration of @xmath238 over the source area gives the total flux in eq .
( [ fsub ] ) . from eq .
( [ undurad5 ] ) and eq .
( [ dfdsforbdivdom ] ) one derives the following relation between on - axis flux density and total flux @xmath371 with this we also point out that in the beam divergence - dominated regime the brightness can be written as @xmath372 a third
interesting limiting case can be considered . in the beam size - dominated regime @xmath373 , eq .
( [ gnor3 ] ) can be simplified as follows .
we can neglect @xmath350 in the exponential functions in @xmath342 .
moreover , the range of variables @xmath374 and @xmath375 is effectively limited up to values of order of unity because they enter the expression for the electric field @xmath290 .
it follows that , when @xmath376 , the exponential functions in @xmath343 can be replaced with unity and we obtain the following expression for the cross - spectral density in dimensional units @xmath377 the wigner distribution in eq .
( [ wig13 ] ) can therefore be written as @xmath378 some simplification may be obtained by rewriting the electric field on the source , @xmath329 , in the terms of the far field @xmath379 . in fact , inserting eq .
( [ farzone ] ) into eq ( [ wthird ] ) , performing the integration and rearranging yields as the integral in eq .
( [ farzone ] ) , after substitution in eq .
( [ wthird ] ) we can present results of integration over @xmath78 and @xmath380 in terms of a dirac @xmath100-function and evaluate all integrals analytically . ] : @xmath381 where @xmath379 is the radiation field from eq .
( [ undurad4bis ] ) .
the peak value of the wigner function is given by @xmath382 where @xmath235 is the maximum of the angular photon flux given by eq .
( [ maxsub ] ) .
generally , calculation of undulator source brightness in the case of partial coherent radiation involves very complicated and time - expensive evaluations .
in fact , one needs to find the maximum of a wigner distribution which is function of four variables ( @xmath383 , @xmath73 ) and depends on four dimensionless problem parameters ( @xmath384 , @xmath385 ) . in some particular cases , however , the brightness can be determined analytically . in section
[ 43 ] we have seen how exact results can be obtained for the diffraction - limited case and in geometrical optics limits .
we now focus our discussion on third - generation sr sources . in this case we can consider @xmath386 and @xmath387 still retaining full generality concerning the values of @xmath388 and @xmath389 , due to the small coupling coefficient between horizontal and vertical emittance .
exploitation of the extra parameters @xmath386 and @xmath387 specializes our theory to the case of third - generation sources . with this in mind
we simplify the wigner distribution calculations beginning with the expression for the cross - spectral density at the virtual source , eq .
( [ gnor3 ] ) .
( [ gnor3 ] ) the range of the integration variable @xmath390 is effectively limited up to values @xmath391 .
in fact , @xmath390 enters the expression for the field @xmath290 .
it follows that at values larger than unity the integrand in eq .
( [ gnor3 ] ) is suppressed .
then , since @xmath386 , we can neglect @xmath390 in the exponential function .
moreover @xmath387 , and from the exponential function in @xmath392 follows that @xmath354 can be neglected in the expression for the field @xmath290 . as a result , eq .
( [ gnor3 ] ) is factorized in the product of horizontal cross - spectral density @xmath393 and a vertical cross - spectral density @xmath394 : @xmath395 where @xmath396 @xmath397 \cr & & \times \widetilde{e}\left(\phi_x , \phi_y + \frac{\delta \hat{y}}{2}\right ) \widetilde{e}^*\left(\phi_x ,
\phi_y -\frac{\delta \hat{y}}{2}\right ) ~. \label{gyfact}\end{aligned}\ ] ] therefore , the wigner distribution in dimensional units for an arbitrary state of coherence in the vertical direction can be written as @xmath398 \widetilde{e}\left ( x ' , y ' + \frac{\delta y}{2}\right)\widetilde{e}^*\left ( x ' , y ' - \frac{\delta y}{2}\right ) \label{wdimv}\end{aligned}\ ] ] it is instructive to check this expression in the case of ( vertical ) beam size- and ( vertical ) divergence - dominated regime . since @xmath399
, one can neglect @xmath89 in the expression for the amplitude of the electric field .
moreover , @xmath400 and the exponential function @xmath401 $ ] is smoothly varying in the region of integration of importance .
therefore , we can take it out of the integral sign , and eq .
( [ wdimv ] ) yields back eq .
( [ wndlargedim ] ) and eq .
( [ bndlarge ] ) as it must be .
some other limiting forms of eq .
( [ wdimv ] ) are of interest , and will be discussed here . in the diffraction - limited regime for the vertical direction one has @xmath402 and @xmath403 .
the distribution @xmath404 $ ] is sharply peaked about @xmath405 , and can be approximated by a dirac @xmath100-function .
moreover , the variable @xmath406 enters the expression for the electric field @xmath290 , and in the region of integration of importance the exponential function in @xmath407 can be replaced by unity .
finally , the field has the symmetry property @xmath408 .
therefore , on - axis , eq .
( [ wdimv ] ) can be simplified to @xmath409 this exact result coincides with the result expected with the help of the approximate formula eq .
( [ bundug ] ) .
let us now consider the beam divergence - dominated regime .
when @xmath410 the variable @xmath89 can be neglected in the expression for the field and the exponential function in @xmath388 , multiplied by @xmath411 , can be approximated by a dirac @xmath100-function under the convolution integral in eq .
( [ gyfact ] ) .
therefore , the expression for the wigner distribution on - axis can be simplified as @xmath412 where @xmath413 is the radiation field in eq .
( [ undurad5 ] ) . accounting for eq .
( [ tflux ] ) the brightness may be written as @xmath414 \frac{1}{4 } \int_{-\infty}^{\infty } d\phi s_0 ^ 2(\phi)~ \label{b2item}\end{aligned}\ ] ] where the dimensionless field distribution @xmath415 is defined by @xmath416 \label{s0}\end{aligned}\ ] ] the expression in the first parenthesis in eq .
( [ b2item ] ) represents the brightness expected by the estimate in eq .
( [ bundug ] ) with the parameter choice in ( [ sigrrp ] ) .
the number @xmath417 represents the disagreement between the exact expression and the estimated maximum of the photon flux density in the phase space .
a second geometrical optics limit for third generation sr sources is the beam size - dominated regime , when @xmath418 .
the wigner distribution eq .
( [ wdimv ] ) can be simplified following the same line of reasoning as for eq .
( [ wthird ] ) . from the analysis of the exponential function in @xmath388 in eq .
( [ gyfact ] ) follows that we can take it out from under the integral over @xmath419 .
moreover , the exponential function in @xmath389 can be replaced by unity and eq .
( [ wdimv ] ) can be written on - axis as @xmath420 if we write the electric field amplitude @xmath413 as the integral in eq .
( [ farzone ] ) , we can present result of the integration over @xmath421,@xmath422 and @xmath89 in eq .
( [ wsizdomax ] ) as dirac @xmath100-functions .
performing the integration and rearranging yields : @xmath423 where the expression for the far field @xmath424 is given in eq .
( [ undurad4bis ] ) . accounting for eq .
( [ maxsub ] ) the brightness can be written as @xmath425\frac{1}{\pi \sqrt{2}}\int_{-\infty}^{\infty } d \phi~ \mathrm{sinc}^2(\phi^2/4)]~ , \label{wsizdomax3}\end{aligned}\ ] ] where @xmath426 is the dimensionless far field distribution .
the factor in square brackets represents the brightness expected by the estimate in eq .
( [ bundug ] ) with the parameter choice in ( [ sigrrp ] ) .
the number @xmath427 = \frac{4}{3}\sqrt{\frac{2}{\pi } } \simeq ~ 1.06 \label{phiphi}\end{aligned}\ ] ] represents , instead , the disagreement between the exact expression and the maximum photon flux density in the phase space .
thus we have found that the exact result in eq .
( [ wsizdomax3 ] ) is naturally different from the estimate in eq .
( [ bundug ] ) , although the difference is not large .
the difference between the value found in eq .
( [ phiphi ] ) , which is close to unity and that in eq .
( [ numberboh ] ) , which is quite different from unity , is striking .
one can observe that this difference is in close relation with other comparisons between the estimate in eq .
( [ bundug ] ) and exact results .
in fact , eq . ( [ bundug ] )
coincides with the exact result eq .
( [ bthirdfin ] ) in the limit for @xmath373 and at the same time was proven to overestimate the exact result by four times in the limit for @xmath428 .
geometry for the analysis of bending magnet brightness.,width=340 ] standard geometry for the analysis of synchrotron radiation from a bending magnet.,width=340 ] consider a single relativistic electron moving on a circular orbit and an observer as sketched in fig .
[ srgeo ] .
it is worth to underline the difference between the geometry depicted in fig .
[ srgeo ] and the geometry used in most sr textbooks for the treatment of bending magnet radiation depicted in fig .
[ srstand ] .
the observer in fig .
[ srstand ] is assumed to be located in a vertical plane tangent to the circular trajectory at the origin , at an angle @xmath248 above the level of the orbit . in other words , in this geometry the @xmath4 axis is not fixed , but depends on the observer s position .
note that the geometry of the electron motion has a cylindrical symmetry , with the vertical axis going through the center of the circular orbit .
because of this symmetry , in order to calculate spectral and angular photon distributions , it is not necessary to consider an observer at a more general location . however , since the wavefront is not spherical , this way of proceeding can hardly help to obtain the phase of the field distribution on a plane perpendicular to a fixed @xmath4 axis .
this is required , for instance , if one needs to calculate the wigner distribution , as in our case .
we can use eq .
( [ generalfin ] ) to calculate the far zone field of radiation from a relativistic electron moving along an arc of a circle . assuming a geometry with a fixed @xmath4 axis as in fig .
[ srgeo ] , we can write the transverse position of the electron as a function of the curvilinear abscissa @xmath429 as @xmath430 and @xmath431 where @xmath195 is the bending radius . since the integral in eq .
( [ generalfin ] ) is performed along @xmath4 we should invert @xmath432 in eq .
( [ zmot ] ) and find the explicit dependence @xmath433 : @xmath434 so that @xmath435 where the expansion in eq .
( [ sz ] ) and eq .
( [ rpdis ] ) is justified , once again , in the framework of the paraxial approximation . with eq .
( [ generalfin ] ) we obtain the radiation field amplitude in the far zone : @xmath436 where @xmath437~.\label{phh2}\end{aligned}\ ] ] one can easily reorganize the terms in eq .
( [ phh2 ] ) to obtain @xmath438~.\label{phh2b}\end{aligned}\ ] ] with redefinition of @xmath439 as @xmath440 under integral we obtain the final result @xcite , @xcite , @xcite : @xmath441}\right\}~,\label{srtwob}\end{aligned}\ ] ] where @xmath442 and @xmath443 in standard treatments of bending magnet radiation , the phase term @xmath444 is absent . in fact , the horizontal observation angle @xmath72 is always equal to zero in the reference system sketched in fig .
[ srstand ] .
the reason for this is that most textbooks focus on the calculation of the intensity radiated by a single electron in the far zone , which involves the square modulus of the field amplitude but do not analyze , for instance , situations like source imaging .
our next goal is to evaluate the integral in eq .
( [ srtwob ] ) .
it is convenient to introduce dimensionless geometrical quantities @xmath445 and the dimensionless parameter @xmath446 if we then go through the algebra we can simplify eq .
( [ srtwob ] ) to @xmath447\exp\left[-\frac{i\hat{\theta}_x}{2}\left(\xi + \frac{\hat{\theta}_x^2}{3}+\hat{\theta}_y^2\right)\right]\cr & & \times\bigg\ { \vec{e}_x \left[(\xi + \hat{\theta}_y^2 ) k_{2/3 } \left(\frac{1}{3 } ( \xi+\hat{\theta}_y^2)^{3/2}\right)\right]\cr & & - i \vec{e}_y \left[(\xi + \hat{\theta}_y^2)^{1/2 } \theta_y k_{1/3}\left(\frac{1}{3 } ( \xi+\hat{\theta}_y^2)^{3/2}\right)\right ] \bigg\}~,\label{ef}\end{aligned}\ ] ] where @xmath448 and @xmath449 are the modified bessel functions .
eq ( [ ef ] ) is equivalent to eq .
( [ srtwob ] ) , but is expressed in a more suitable form for calculating the field distribution at the virtual source , which is assumed to be located in the @xmath98-plane , perpendicular to the circular trajectory at the origin ( see fig .
[ srgeo ] ) .
after substitution of eq .
( [ ef ] ) in eq .
( [ farzone ] ) we can write result in terms of the airy functions @xmath450 ( \xi+\hat{\theta}_y^2 ) k_{2/3 } \left[\frac{1}{3 } ( \xi + \hat{\theta}_y^2)^{3/2}\right ] \mathrm{ai}\left[\frac{1}{2^{2/3 } } ( \xi - 2 \hat{x } + \hat{\theta}_y^2)\right ] \cr & & \label{e0x}\end{aligned}\ ] ] and @xmath451 k_{1/3 } \left[\frac{1}{3 } ( \xi + \hat{\theta}_y^2)^{3/2}\right ] \mathrm{ai}\left[\frac{1}{2^{2/3 } } ( \xi - 2 \hat{x } + \hat{\theta}_y^2)\right]~. \label{e0y}\end{aligned}\ ] ] , as a function of @xmath452 and @xmath453 for the horizontal ( left plot ) and vertical ( right plot ) polarization components.,title="fig:",scaledwidth=50.0% ] , as a function of @xmath452 and @xmath453 for the horizontal ( left plot ) and vertical ( right plot ) polarization components.,title="fig:",scaledwidth=50.0% ] the integral in eq .
( [ e0x ] ) and eq .
( [ e0y ] ) must be calculated numerically .
computational results of the intensity distribution of horizontal and vertical polarization components in the 1:1 image plane by perfect thin lens @xcite are presented in fig.[vsxy ] .
due to phase differences of the bending magnet radiation in the far zone from a spherical wavefront , the wavefront is aberrated . for a single electron
this aberration appears when the horizontal aperture is compatible with the natural opening angle @xmath178 , and it becomes severe as the horizontal aperture increases further . up to this point we considered an electron moving along a circular trajectory that lies in the @xmath454-plane and following a path through origin of the coordinate system sketched in fig .
[ srgeo ] , tangent to the @xmath4 axis .
the phase difference in the fields will be determined by the position of the observer position and by the electron trajectory .
let us now discuss the bending magnet radiation from a single electron with arbitrary angular deflection and offset with respect to the nominal orbit .
such an expression was first calculated , starting from the lienard - wiechert fields , in @xcite .
the meaning of horizontal and vertical deflection angles @xmath455 and @xmath456 is clear once we specify the electron velocity @xmath457~,\cr & & \label{veloangle}\end{aligned}\ ] ] so that the trajectory can be expressed as a function of the curvilinear abscissa @xmath429 as @xmath458 \vec{e_x } \cr & & + \left[l_y+s \sin(\eta_y)\right ] \vec{e_y } \cr & & + \left [ r\sin\left({s\over{r}}+\eta_x\right)\cos(\eta_y ) - r\sin(\eta_x)\sin(\eta_y ) \right ] \vec{e_z } \cr & & \label{trajangle}\end{aligned}\ ] ] here we have introduced , also , an arbitrary offset @xmath459 in the trajectory . using eq .
( [ trajangle ] ) an approximated expression for @xmath433 can be found : @xmath460 so that @xmath461 and @xmath462 it is evident that the offsets @xmath463 and @xmath464 are always subtracted from @xmath465 and @xmath466 respectively : a shift in the particle trajectory on the vertical plane is equivalent to a shift of the observer in the opposite direction . with this in mind
we introduce angles @xmath467 and @xmath468 to obtain @xmath469 and @xmath470 one can easily reorganize the terms in eq .
( [ phh2ang ] ) to obtain @xmath471 redefinition of @xmath439 as @xmath472 gives the result @xmath473}\right\}~,\label{srtwoang2}\end{aligned}\ ] ] where @xmath474 and @xmath475 in the far zone we can neglect terms in @xmath476 and @xmath477 , which leads to @xmath478}\right\}~,\label{srtwoang2bis}\end{aligned}\ ] ] where @xmath479 and @xmath480 note that eq .
( [ srtwoang2bis ] ) and eq .
( [ srtwob ] ) satisfy the equality in eq .
( [ tiltshiftfar ] ) . therefore , also in the bending magnet case the statistical average can be simplified to the convolution integral eq .
( [ wcorr ] ) .
we first calculate the brightness for a filament beam and , as a second step , we account for a finite phase space distribution for the electron beam .
the wigner distribution in the case of an electron beam with zero emittance is given by eq .
( [ w0start ] ) , where @xmath329 is related to the electric field radiated by a single electron in a bending magnet , eq .
( [ e0x ] ) , where we now restrict our attention to the horizontal polarization component . according to eq .
( [ w0start ] ) and eq .
( [ e0x ] ) , the wigner distribution in the diffraction - limited case is a function of four geometrical variables @xmath481 , @xmath482 , @xmath483 , @xmath484 and one parameter @xmath485 . we can compute the bending magnet brightness numerically , as a function of these variables .
the peak value of the wigner distribution at @xmath486 as a function of @xmath487 and @xmath488 is reached for @xmath489 , @xmath490 , @xmath491 and @xmath492 .
therefore , the expression for the brightness is given by @xmath493 \label{bbm1}\end{aligned}\ ] ] and , according to the usual normalization , it yields a number of photons per relative bandwidth per unit time per unit area per unit solid angle .
apart from a numerical factor , it equals the theoretical maximum concentration of the photon flux on the sample on the basis of qualitative arguments .
the number of photons with @xmath494 emitted from a formation length @xmath495 per unit time can be estimated as @xmath496 .
this radiation concentrates around the axis within a solid angle of about @xmath213 and , as a result , it has an @xmath212 factor close to unity . in literature
it is often noted that the coherent flux from radiation with an ideal wavefront can be ultimately focused down to a spot - size of dimension @xmath176 .
it is clear that the numerical factor @xmath497 is related with the particular choice of wavelength , and will be different for different choices of @xmath498 .
the analysis of the effects of a finite electron beam emittance can be made using the same approach as in section [ sec : due ] .
the calculations are easier if we make use of dimensionless variables and parameters for the electron beam distribution .
normalized units in the bending magnet case are given by eq .
( [ normqr ] ) and its analogues @xmath499 as for the undulator case , in order to make our analysis treatable we assume a gaussian distribution of the electron beam in the transverse phase space , which can be factorized at the position of the virtual source ( at @xmath338 ) . in this case
, the electron beam distribution at the source position can be expressed as in eq .
( [ fphase ] ) and eq .
( [ distr ] ) . in the bending magnet case ,
the dimensionless parameters @xmath342 and @xmath343 are given by @xmath500 we are now in the position to calculate the cross - spectral density at the virtual source by taking advantage of our dimensionless analysis .
the expression for the cross - spectral density @xmath92 is in eq .
( [ gnor3 ] ) , where now dimensionless variables and parameters are given by eq .
( [ normqr ] ) , eq .
( [ normq1 ] ) and eq .
( [ nd2 ] ) . the expression for the two polarization components of the electric field , eq .
( [ e0x ] ) and eq .
( [ e0y ] ) , allow for an explicit calculation of @xmath501 .
the final step consists in the calculation of the wigner distribution @xmath96 .
the relation between @xmath96 and @xmath92 is expressed by the transformation in eq .
( [ wig13 ] ) , so that @xmath502 the integral in eq .
( [ wgdless ] ) is a function of four dimensionless variables @xmath503 , and five dimensionless parameters @xmath342 , @xmath343 , and @xmath498 . in order to calculate the brightness we need to find the maximum of the wigner distribution .
suppose that the maximum of @xmath96 is reached at @xmath504 and @xmath505 .
then one has @xmath506 where @xmath37 is the integral in eq .
( [ wgdless ] ) . here
we have , at last , a well - defined procedure for computing the brightness from a bending magnet source .
thus , in principle , we have solved the problem of determining the brightness of a given sr setup . in particular , in the geometrical optics limits , when parameters @xmath342 or @xmath343 are large , calculations become simple and it is possible to calculate the brightness analytically .
when the parameters @xmath342 and @xmath343 are order of unity the situation becomes more complicated , and must be solved numerically .
we will not further pursue the matter of determining the brightness with the help of numerical techniques . for the discussion of the next section [ geosrlim ] attention
will be restricted to the geometrical optics limit only .
it is our purpose here to demonstrate how a straightforward application of eq .
( [ balgo ] ) yields analytical expressions for the brightness in several geometrical optics limits .
we begin our analysis of geometrical optics asymptotes with the beam divergence - dominated regime , for @xmath428 , and with the beam size - dominated regime for @xmath373 . before proceeding , however
, we should first make a few remarks to discuss the choice of these examples .
in fact , in section [ 43 ] we considered the beam size- and divergence - dominated regime as the simplest geometrical optics asymptote for the undulator case .
in contrast with this , in the bending magnet case the beam size- and divergence - dominated regime , corresponding to @xmath352 and @xmath353 , is the most complicated situation for analytical treatment . on intuitive grounds we certainly expect that in this asymptotic limit there is a competition between effects related with partial coherence ( since @xmath352 ) and what we called ` aberration ' effects ( since @xmath353 ) , and careful mathematical analysis confirms such expectation . given its complexity , we will discuss this asymptote only as a final example . from a mathematical viewpoint
there are two differences between the bending magnet case and the undulator case .
the first difference stems from the fact that the range of the variables @xmath350 in eq .
( [ gnor3 ] ) is effectively limited up to values @xmath351 in the undulator case .
in fact , @xmath350 enters the expression for electric field eq .
( [ undurad5 ] ) , and at values larger than unity the integrand in eq .
( [ gnor3 ] ) is suppressed .
therefore , when @xmath352 we can neglect @xmath350 in the exponential functions under the integral . at variance , in the bending magnet case the expression for the field is given by eq .
( [ e0x ] ) and eq .
( [ e0y ] ) , and the range of the variable @xmath390 in eq .
( [ gnor3 ] ) is not limited .
therefore , eq . ( [ gnor3 ] ) can not be simplified by neglecting @xmath390 in the exponential function in @xmath507 under the integral sign only based on the assumption @xmath386 .
the second difference can be found in the fact that , in the undulator case for @xmath353 , from the exponents in @xmath343 in eq .
( [ gnor3 ] ) follows that @xmath508 and @xmath509 , and can be neglected in the expression for the field .
in contrast with this , in the bending magnet case , from the expression for the electric field on the source position ( eq . ( [ e0x ] ) and eq . ( [ e0y ] ) ) it is clear that there are strong oscillations with a decreasing ` local period ' along the @xmath510 direction .
thus , it can be seen that @xmath88 can not be ignored at @xmath387 , and the integral in eq .
( [ gnor3 ] ) can not be simplified to an integral of the flux density over the source surface , as was done for the undulator case , eq .
( [ wndlargedim ] ) .
let us now consider the beam divergence - dominated regime for @xmath428 . from the analysis of the exponential functions in @xmath343 in eq .
( [ gnor3 ] ) follows that these functions have a significant magnitude only when @xmath511 and @xmath512 .
then , since @xmath362 , it follows that the distributions @xmath513 $ ] and @xmath514 $ ] are sharply peaked about @xmath515 and @xmath366 , respectively .
the integrals along the @xmath510 axis and the @xmath516 axis in eq .
( [ gnor3 ] ) , including the two polarization components for the field @xmath517 given by eq .
( [ e0x ] ) and eq .
( [ e0y ] ) , can therefore be simplified by the same line of thought as for the undulator case .
however , the important difference with the undulator case is that now the inequality @xmath518 can not be called upon for ignoring @xmath374 , thus approximating the exponential function in @xmath507 , multiplied by @xmath519 , with a dirac @xmath100-function under the integral sign at any point @xmath481 and @xmath482 of the source surface .
in fact , based on the expressions for the electric field amplitude at the source position , eq .
( [ e0x ] ) and eq .
( [ e0y ] ) , we might argue that the exponential function in @xmath507 , multiplied by @xmath519 , can be approximated by a dirac @xmath100-function and @xmath374 can be neglected only if @xmath520 and @xmath521 are small compared to the ` local period ' of oscillation of the field along the @xmath510 axis .
this requirement ( at fixed @xmath507 , @xmath392 ) would set an upper limit on the @xmath510-coordinate when the integrand can be simplified as described above .
the situation can be summarized by saying that the asymptotic expression for the wigner distribution in the beam divergence - dominated regime in eq .
( [ wdivdom ] ) can be used in the bending magnet case at least in the region where @xmath522 and @xmath523 .
this is in contrast with the undulator case , when eq .
( [ wdivdom ] ) can be used at any point on the source plane . according to eq .
( [ e0x ] ) , for the horizontal polarization component in the diffraction - limited case , the peak value of the square modulus of the electric field amplitude on the source at @xmath524 is reached at @xmath490 and @xmath525 ( see fig . [ vsxy ] ) .
this point is well within the region of applicability of eq .
( [ wdivdom ] ) for @xmath428 .
thus , the expression for the brightness of the horizontal polarization component is given by ) for the undulator case is striking here . ] : @xmath526 where @xmath527 is the maximum photon flux density for the horizontal polarization component of the field of the bending magnet source in the diffraction - limited case . for @xmath217 eq .
( [ bxbend ] ) can also be written as @xmath528 ~ , \label{bmaxfluxd}\end{aligned}\ ] ] where @xmath178 is the so - called ` vertical opening angle ' @xcite introduced already in section [ oldres ] . at
@xmath194 one finds , for the horizontal polarization component , that @xmath193 .
the numerical factor is related with the particular choice of wavelength , and will be different for different choices of @xmath498 , while the parametric dependence in square brackets remains unchanged . a second , simple limiting case can now be considered . in the beam size - dominated regime ,
when @xmath373 , eq .
( [ gnor3 ] ) can be simplified following the same line of reasoning as in the case of an undulator .
the only difference is in the region of applicability of this simplification .
we start asserting the simplifications that can be made and examining their implications .
variables @xmath374 and @xmath375 in eq .
( [ gnor3 ] ) are effectively limited up to values @xmath529 and @xmath530 , if we limit the observation angles up to @xmath531 , @xmath532 .
therefore , the exponential functions in @xmath343 in eq .
( [ gnor3 ] ) can be replaced with unity .
moreover @xmath352 , so that we can neglect @xmath350 in the exponential function in @xmath342 . the wigner distribution can therefore be written as eq .
( [ wthird ] ) .
the integral in this equation can be simplified by rewriting the electric field at the source , eq .
( [ e0x ] ) and eq .
( [ e0y ] ) , in the terms of the far field given by eq .
( [ ef ] ) .
in fact , inserting eq .
( [ tiltshiftfar ] ) into eq .
( [ wthird ] ) , performing integration and rearranging yields eq .
( [ wthirdfin ] ) , where now the electric field amplitude in the far zone is given by eq .
( [ ef ] ) . a formal way to understand
this simplification procedure is the following .
the wigner distribution @xmath96 is proportional to the two - dimensional fourier transform of the cross - spectral density @xmath92 , and the coordinates @xmath533 can be interpreted as reduced spatial frequencies .
suppose now that we have @xmath531 .
the significant values of @xmath374 in the integral correspond , then , to an argument of the electric field amplitude @xmath290 of order of unity .
this is because the reduced ` local period ' of the electric field amplitude oscillations is of order of unity near the origin and decreases along the @xmath510 axis . in the case
when spatial frequencies are of order unity , the main contribution to the integral comes from the region near @xmath534 and @xmath535 , and its width is of order @xmath529 , @xmath391 . since the function @xmath536 $ ]
is smoothly varying in this region we can replace it by @xmath537 $ ] , and take it out the integral sign .
we can also replace the exponential function in @xmath343 with unity , and the integral is then approximately given by eq .
( [ wthird ] ) . up to now , for our calculations we used the wigner distribution method at the virtual source position .
from the discussion concerning the beam size - dominated regime for bending magnet sources , it is obvious that such choice is not convenient in this asymptotic .
in contrast to the beam divergence - dominated regime , the approximation procedure described above is not straightforward , and involves a number of subtleties .
calculations of @xmath96 can be performed in two different ways , both consistently leading to the same result , and a comparison between the different calculation procedures is instructive .
the approximation procedure in the beam size - dominated regime is easily understood if , instead of the virtual source , one uses the far zone for calculating @xmath96 .
this approach also offers an opportunity to easily find the condition for applicability of the approximation made .
as explained in section [ sec : wiggen ] , the standard way of deriving the expression for the wigner distribution makes use of the expression for the cross - spectral density @xmath92 . in this way
, our problem is reduced to the calculation of the cross - spectral density in the far zone .
it can be interesting to explore the properties of the cross - spectral density and see that an explicit expression in the far zone can be derived from eq .
( [ gnor3 ] ) , describing @xmath92 at the source position , by symmetry under the duality transformations : @xmath538 the cross - spectral density @xmath92 in the far zone is therefore given by @xmath539 \exp \left[-\frac{(\delta \hat{\theta}_y)^2 n_y}{2}\right]\cr & & \times \int_{-\infty}^{\infty } d \phi_x \int_{-\infty}^{\infty } d \phi_y \exp\left[-\frac{\left(\phi_x-\hat{\theta}_x\right)^2}{2 d_x}\right ] \exp\left[-\frac{\left(\phi_y-\hat{\theta}_y\right)^2}{2 d_y}\right ] \cr & & \times \widetilde{e}\left({\phi_{x}}+\frac{\delta \hat{\theta}_x}{2},{\phi_{y}}+\frac{\delta \hat{\theta}_y}{2 } \right ) \widetilde{e}^*\left({\phi_{x}}-\frac{\delta \hat{\theta}_x}{2},{\phi_{y}}-\frac{\delta \hat{\theta}_y}{2 } \right)~,\cr & & \label{gnor3far}\end{aligned}\ ] ] where the amplitude of the electric field @xmath540 in the far zone is given by eq .
( [ ef ] ) .
note that the mathematical rule that we just gave to calculate the cross - spectral density in the far zone differs from the usual approach , where the angular and the position representations of the field , @xmath329 and @xmath379 are related via a spatial fourier transform and the fields are given in the form of angular spectrum of plane waves .
in that way , the statistical properties of an ensemble of fields at the source position is fully reflected in the ensemble of the angular spectrum amplitudes , and the cross - spectral density is a measure of the correlation between the fields of the plane - wave modes propagating at @xmath541 and @xmath542 . however , the correlation between plane waves is a purely mathematical quantity that depends , for example , on the normalization convention used .
in contrast with this , in our approach we do not make any use of spatial fourier transforms , because the angular and the position representations of the field are related via the fresnel propagation eq .
( [ farzone ] ) and eq .
( [ farzonef ] ) .
the field propagation in free space is a physical process , and does not depend on the definition of the spatial fourier transform .
we now return to our quantitative discussion concerning the beam size - dominated example .
if @xmath376 , we can replace the exponential functions in @xmath343 multiplied by @xmath543 with dirac @xmath100-functions , so that the expression for @xmath92 in the far zone , eq .
( [ gnor3far ] ) , can be approximated with @xmath544 \exp \left[-\frac{(\delta \hat{\theta}_y)^2 n_y}{2}\right ] \cr & & \times \widetilde{e } \left(\hat{\theta}_x + \frac{\delta \hat{\theta}_x}{2 } , \hat{\theta}_y + \frac{\delta \hat{\theta}_y}{2}\right ) \widetilde{e}^*\left(\hat{\theta}_x - \frac{\delta \hat{\theta}_x}{2 } , \hat{\theta}_y - \frac{\delta \hat{\theta}_y}{2}\right ) ~. \label{gfarappr}\end{aligned}\ ] ] in order to proceed it is necessary to determine the behavior of the @xmath545 along the horizontal axis .
notice that in the far zone , according to eq .
( [ ef ] ) , the module of the electric field does not depend on @xmath483 , while its phase varies as @xmath546 at @xmath547 . consequently , eq . ( [ gfarappr ] )
depends on @xmath548 only through the phase , and one can further approximate the phase as @xmath549 . if @xmath386 , because of the exponential function @xmath550 in eq .
( [ gfarappr ] ) , the region of @xmath548 for which @xmath92 is large is near the point @xmath551 , and its width is of order @xmath552 .
suppose we have @xmath553 .
we can then approximate @xmath554 , and consequently the phase factor in eq .
( [ ef ] ) is about equal to unity .
we thus obtain @xmath555 \cr & & \times |\widetilde{e}({\theta}_x , { \theta}_y)|^2~. \label{gfarappr2}\end{aligned}\ ] ] in the geometrical optics limit , the cross - spectral density at the source position can be written in the form of eq .
( [ qh ] ) , meaning that variables @xmath5 and @xmath78 are separable . however , according to eq .
( [ gfarappr2 ] ) , in the far zone there is no separation of variables @xmath97 and @xmath556 in @xmath92 , even in the geometrical optics limit .
this is contrast with the traditional definition of cross - spectral density in the far zone as a correlation function between plane - wave modes .
the non - separability in eq .
( [ gfarappr2 ] ) is a consequence of the fact that in our definition of cross - spectral density , eq . ( [ gnor3far ] ) , we use the complete field amplitudes in the far zone , which include a spherical phase factor . the condition for the applicability of eq .
( [ gfarappr2 ] ) is @xmath553 for @xmath373 . for our purposes
it is preferable to express the wigner distribution at the source position in terms of the cross - spectral density in the far - zone .
the field at @xmath338 is related to the field in the far zone by eq .
( [ farzonef ] ) . inserting this expression into eq .
( [ wig13 ] ) , one finds @xmath557 if we now substitute eq .
( [ gfarappr2 ] ) in eq .
( [ wigfarfin ] ) we obtain eq .
( [ wthirdfin ] ) as must be , but with the alternate derivation above we specified more precisely the conditions of applicability of eq .
( [ wthirdfin ] ) .
let us finally investigate the more complicated beam size- and divergence - dominated regime .
formally we shall consider the limiting case when the dimensionless parameters of the electron beam distribution @xmath342 and @xmath343 are much larger than unity .
however , even at @xmath353 and @xmath352 we have to distinguish between two limiting expressions for the brightness of a bending magnet , at variance with the single result obtained for undulators .
in fact , for bending magnets , the dimensionless parameter @xmath558 plays an important role , and one should additionally consider two limiting cases for @xmath558 much smaller or much larger than unity . to explore the nature of this extra parameter
, we begin by noting that another situation where a similar parameter exists was already discussed above , when treating the condition of applicability for the approximation of the wigner distribution in the beam size - dominated regime . from that case
one can show that the condition @xmath554 is a consequence of the small ` aberration ' influence within a window centered in the far zone at an angle @xmath483 with a horizontal opening angle @xmath559 .
in fact , the electron motion in a bend has cylindrical symmetry with the vertical axis going through the center of the circular orbit .
therefore , an observer on - axis ( fig .
[ srstand ] ) receives as much radiation from an electron with horizontal deflection angle @xmath483 as an observer looking at an electron with zero deflection angle from an angle @xmath483 .
it follows that the existence of the parameter @xmath558 is a consequence of the small ` aberration ' influence within a window placed in the far zone about the @xmath4-axis , with a horizontal opening angle @xmath560 .
finally , it is worth emphasizing that the horizontal opening angle @xmath552 is the angular dimension of a coherent area in the far zone .
therefore , one can summarize the previous observations by saying that the sr emitted within a solid angle of about @xmath561 is fully transversely coherent and additionally , at @xmath562 has an @xmath212 factor close to unity .
such fraction of radiation can be ultimately focused down to a spot size of dimension of about @xmath225 . in light of this
, one sees that @xmath558 is an important problem parameter that is required in order to calculate the bending magnet brightness in the beam size- and divergence - dominated regime .
in particular , based on this intuitive reasoning we expect that in the asymptotic limit eq .
( [ condm2 ] ) the brightness should be proportional to the coherent solid angle @xmath563 , but should not depend on the electron beam divergence in the horizontal direction , even when @xmath387 . in the following
we will demonstrate that rigorous mathematical analysis confirms such expectation .
we begin by approximating the expression for the cross - spectral density in the far zone , eq .
( [ gnor3far ] ) , in the limiting case for @xmath353 and @xmath352 .
suppose that condition eq .
( [ condm2 ] ) is satisfied . from the analysis of the exponential functions in @xmath342 in eq .
( [ gnor3far ] ) follows that @xmath559 and @xmath564 can be neglected in the expression for the field .
thus , the cross - spectral density is a product of two separate factors .
the first is the fourier transform of the transverse electron beam distribution .
the second is the convolution of the electron beam angular distribution with the two dimensional function @xmath565 , which is proportional to the angular intensity distribution of the radiation .
moreover , the range of the variable @xmath419 in this convolution integral is effectively limited up to values @xmath566 .
in fact , @xmath419 enters the expression for the modulus of the electric field ( see eq .
( [ ef ] ) ) , and at values larger than unity the integrand in the convolution is suppressed .
it follows that when @xmath399 we can neglect @xmath419 in the exponential function @xmath567 $ ] under the integral sign .
( [ gnor3far ] ) is therefore approximated by @xmath568 |\widetilde{e}(\phi_x , \phi_y)|^2~. \cr & & \label{gnor3farappr}\end{aligned}\ ] ] since @xmath569 does not depend on @xmath390 , we can take it out from under the integral over @xmath390 .
if we now substitute this expression in the definition of the wigner distribution , eq .
( [ wigfarfin ] ) and perform the integral over @xmath570 we obtain in dimensional units @xmath571 where @xmath572 is the photon flux per unit time per unit horizontal angle per relative spectral bandwidth .
the brightness approximated by eq ( [ btext ] ) coincides with the maximum of @xmath96 in eq .
( [ wfinfin ] ) in the case when @xmath353 and @xmath352 and at arbitrary electron beam divergence in horizontal direction ( i.e. at arbitrary @xmath392 ) .
now we demonstrated that the condition for validity of eq .
( [ wfinfin ] ) is eq .
( [ condm2 ] ) .
let us now consider the asymptotic case opposite to ( [ condm2 ] ) : @xmath573 up to now , for our calculations in the beam size- and divergence - dominated regime ( @xmath352 and @xmath353 ) we used the wigner function method in the far zone . in the asymptotic case ( [ condm2poor ] ) it is more convenient to use the virtual source for calculating @xmath96 .
the cross - spectral density at the virtual source position is given by eq .
( [ gnor3 ] ) . from the analysis of the exponential function in @xmath343 in eq .
( [ gnor3 ] ) follows that @xmath374 and @xmath375 can be neglected in the expression for the field .
thus , the cross - spectral density is a product of two separate factors .
the first is the fourier transform of the angular electron beam distribution .
the second is the convolution of the electron beam spatial distribution with the two dimensional function @xmath565 , which is proportional to the intensity distribution of the radiation from a single electron at the source position .
the range of the variable @xmath419 in the convolution integral is effectively limited up to values @xmath566 , because @xmath419 enters the expression for module of electric field eq .
( [ e0x ] ) . therefore ,
when @xmath400 we can neglect @xmath419 in the exponential function @xmath574 $ ] under the convolution integral . the approximation of eq .
( [ gnor3 ] ) is then given by @xmath575\exp\left[-\frac{(\delta \hat{y})^2 d_y } { 2}\right ] \exp\left(-\frac{\hat{y}^2}{2 n_y}\right ) \cr & & \times \int_{-\infty}^{\infty } d \phi_x \int_{-\infty}^{\infty } d
\phi_y \exp\left[-\frac{(\phi_x-\hat{x})^2}{2 n_x}\right ] |\widetilde{e}(\phi_x , \phi_y)|^2 ~ , \label{gapprlast}\end{aligned}\ ] ] where @xmath290 is the radiation field in eq .
( [ e0x ] ) .
since @xmath576 depends on @xmath390 , we can not take it out from under the integral .
if we substitute this approximation of @xmath92 in the definition of the wigner function eq .
( [ wgdless ] ) and perform the prescribed integration we obtain the following result in dimensional units @xmath577 \label{wfin5p3}\end{aligned}\ ] ] comparing eq .
( [ gapprlast ] ) and eq .
( [ wfin5p3 ] ) one can see that in the beam size- and divergence - dominated regime , depending on the specific ratio between horizontal beam size and divergence , the wigner distribution is described using functions with completely different parametric dependence .
the brightness approximated by eq .
( [ btext ] ) does not coincide with the maximum of @xmath96 in the case when condition ( [ condm2poor ] ) is satisfied .
finally , we should note that the usually accepted approximation for bending magnet brightness turns out to be parametrically inconsistent not only in the intermediate geometrical optics asymptote when the beam divergence dominates over the diffraction angle , but also in the ` simplest ' geometrical optics asymptotic case when both beam size and beam divergence dominate over diffraction size and diffraction angle .
this document discusses the relation between statistical optics and the electromagnetic theory of sr radiation . a basic problem pertaining this relation is the definition of the sr brightness in terms of electromagnetic fields and their statistical properties .
we consider the brightness defined with the help of a wigner distribution .
we propose a mathematical scheme and formulate a novel theory of sr brightness defined as the maximum of the wigner distribution .
formulating the theory of brightness in the language of wigner distributions has only one guideline , a particular correspondence principle .
the conceptual foundation of this correspondence principle is based on the assumption that the formalism involved in the calculation of brightness must include radiometry as a limiting case .
we use the classical definition of radiance to obtain a correct proportionality factor in the definition of brightness . in this way , in the geometrical optics limit , the brightness can be represented as the maximum value of the radiance . in classical radiometry
the maximum of the radiance is the maximum of the photon flux density in phase space .
since the correspondence principle is integrated once and for all in the very foundation of our theory , one expects to eliminate the necessity of its recurrent application to every individual problem .
we compute various geometrical optics limits according to our definition of brightness , and we compare results with expectations from theories currently used in literature . in many cases we find a significant disagreement between exact calculations of the maximum photon flux density in phase space and the usually accepted approximations for undulator and bending magnet brightness .
we thank oleg gorobtsov for carefully reading our manuscript and for useful discussions . 99 g. green `` spectra and optics of synchrotron radiation '' , bnl-50522 , april 1976 .
s. krinsky , m. perlman and r. watson , `` characteristics of synchrotron radiation and its sources '' in handbook on synchrotron radiation vol 1 , edited by e. koch , north holland publishing company ( 1983 ) .
k. kim , nucl .
instrum . and methods in phys .
res . a 246 ( 1986 ) 71 .
k. kim , nucl .
instrum . and methods in phys .
res . a 261 ( 1987 ) 44 .
k. kim , proceedings of pac 1997 , p. 194
s. hulbert and j. weber , nucl .
instrum and methods a319 25 - 31 ( 1992 ) m. howells and b. kincaid , `` the properties of undulator radiation '' , in a. schlachter and f. wulleumeir ( eds ) , new directions in research with third - generaion soft - x - ray synchrotron radiation sources , 315 - 358 , kluwer academic publishers ( 1994 ) .
s. l. hulbert and g. p. williams , experimetal methods in the physical sciences , vol .
31 , 1 ( 1998 ) s. hulbert and g. williams `` synchrotron radiation sources '' , bnl-65801 , july 1998 .
r. bosch nim a 431 ( 1999 ) 320 - 333 r. bosch nim a 454 ( 2000)497 - 505 g williams , rep
69 ( 2006 ) 301 - 326 i. bazarov , prstab 15 , 050703 ( 2012 ) .
z. huang , slac - pub-15449 and proceedings of the fel conference 2013 , new york ( 2013 ) .
d. attwood , soft x - rays and extreme ultraviolet radiation ( cambridge university press , 1999 ) f. ciocci , g. dattoli , a. torre and a. renieri , insertion devices for synchrotron radiation and free electron laser ( world scientific , 2000 ) p. duke , synchrotron radiation ( oxford university press , 2000 ) h. wiedemann , synchrotron radiation ( springer , 2002 ) h. onuki and p. elleaume , undulators , wigglers and their applications ( taylor @xmath578 francis , 2003 ) a. hofmann , the physics of synchrotron radiation cambridge monographs on particle physics , nuclear physics and cosmology 20 isbn 0521 30826 7 ( 2004 ) .
j. clarke , the science and technology of undulators and wigglers ( oxford university press , 2004 ) r. talman , accelerator x - ray sources ( wiley - vch verlag gmbh , 2006 ) a. m. pravilov , `` radiometry in modern scientific experiments '' springer wien new - york ( 2011 ) j. palmer , b. grant , `` the art of radiometry '' spie press bellingham , washington usa ( 2010 ) w. wigner , phys . rev 40 , 749 ( 1932 ) .
j. goodman , `` statistical optics '' , john wiley and sons inc , ( 2000 ) .
l. mandel abd e. wolf , `` optical coherence and quantum optics '' , cambridge university press ( 1995 ) j. jackson , `` classical electrodynamics '' , 3rd ed .
, wiley , new york ( 1999 ) a.e .
siegman , how to ( maybe ) measure laser beam quality , diode pumped solid state lasers : applications and issues , isbn : 1 - 55752 - 520-x ( 1998 ) g. geloni , e. saldin , e. schneidmiller and m. yurkov , opt .
276 , 167 ( 2007 ) o. chubar , p. elleaume , proceedings of epac-98 , 1777 ( 1998 ) n. smolyakov , proceedings of epac 1998 , p. 1601
( 1998 ) y. takayama and s. kamada , phys .
e 59 , ( 1999 ) o. chubar , p. elleaume and a. snigirev , nim a 435 ( 1999)495 - 508
calculations pertaining undulator radiation are well established .
see e.g. @xcite . in this appendix we present a simple derivation of the frequency representation of the radiated field produced by an electron in an undulator . for the electron transverse velocity we assume usually
, it does not make sense to calculate the intensity distribution from eq .
( [ undurad_ap1 ] ) alone , without extra - terms ( both interfering and not ) from the other parts of the electron trajectory .
this means that one should have complete information about the electron trajectory and calculate extra - terms to be added to eq .
( [ undurad_ap1 ] ) in order to have the total field from a given setup .
yet , we can find _ particular situations _ for which the contribution from eq .
( [ undurad_ap1 ] ) is dominant with respect to others . in this case eq .
( [ undurad_ap1 ] ) , alone , has independent physical meaning .
one of these situations is when the resonance approximation is valid .
this approximation does not replace the paraxial one , based on @xmath589 , but it is used together with it .
it takes advantage of another parameter that is usually large , i.e. the number of undulator periods @xmath321 . in this case , the integral in @xmath590 in eq .
( [ undurad_ap1 ] ) exhibits simplifications , independently of the frequency of interest due to the long integration range with respect to the scale of the undulator period .
@xmath591 \frac{i \omega e}{c^2 z_0 } \cr & & \times \int_{-l/2}^{l/2 } dz'\left\{\frac{k}{2
i \gamma}\left[\exp\left(2 i k_w z'\right)-1\right]\vec{e}_x + \vec{\theta}\exp\left(i k_w z'\right)\right\ } \cr & & \times \exp\left[i \left(c + { \omega \theta^2 \over{2 c } } \right ) z ' - { k\theta_x\over{\gamma}}{\omega\over{k_w c}}\cos(k_w z ' ) - { k^2\over{8\gamma^2 } } { \omega\over{k_w c } } \sin(2 k_w z ' ) \right ] ~. \cr & & \label{undurad2_ap1}\end{aligned}\ ] ] @xmath591 \frac{i \omega e}{c^2 z_0 } \sum_{m , n=-\infty}^\infty j_m(u ) j_n(v ) \exp\left[\frac{i \pi n}{2}\right ] \cr & & \times \int_{-l/2}^{l/2 } dz'\exp\left[i \left(c + { \omega \theta^2 \over{2 c } } \right ) z'\right ] \cr & & \times \left\{\frac{k}{2 i \gamma } \left[\exp\left(2 i k_w z'\right)-1\right]\vec{e}_x + \vec{\theta}\exp\left(i k_w z'\right)\right\ } \exp\left[i ( n+2 m ) k_w z'\right ] ~,\cr & & \label{undurad3_ap1}\end{aligned}\ ] ] up to now we just re - wrote eq .
( [ undurad_ap1 ] ) in a different way .
( [ undurad_ap1 ] ) and eq .
( [ undurad3_ap1 ] ) are equivalent .
of course , definition of @xmath599 is suited to investigate frequencies around the fundamental harmonic but no approximation is taken besides the paraxial approximation .
the first phase term in @xmath439 under the integral sign in eq . ( [ undurad3_ap1 ] ) is varying slowly on the scale of the undulator period @xmath242 . as a result , simplifications arise when @xmath601 , because fast oscillating terms in powers of @xmath602 $ ] effectively average to zero .
when these simplifications are taken , resonance approximation is applied , in the sense that one exploits the large parameter @xmath321 .
this is possible under condition ( [ eqq_ap1 ] ) .
note that ( [ eqq_ap1 ] ) restricts the range of frequencies for positive values of @xmath599 independently of the observation angle @xmath603 , but for any value @xmath604
( i.e. for wavelengths longer than @xmath605 ) there is always some range of @xmath248 such that eq .
( [ eqq_ap1 ] ) can be applied .
altogether , application of the resonance approximation is possible for frequencies around @xmath606 and lower than @xmath606 . once any frequency is fixed , ( [ eqq_ap1 ] ) poses constraints on the observation region where the resonance approximation applies .
similar reasonings can be done for frequencies around higher harmonics with a more convenient definition of the detuning parameter @xmath599 .
note that this condition on frequencies automatically selects observation angles of interest @xmath608 .
in fact , if one considers observation angles outside the range @xmath608 , condition ( [ eqq_ap1 ] ) is not fulfilled , and the integrand in eq .
( [ undurad3_ap1 ] ) exhibits fast oscillations on the integration scale @xmath165 . as a result ,
one obtains zero transverse field , @xmath609 , with accuracy @xmath317 . under the constraint imposed by ( [ resext_ap1 ] ) , independently of the value of @xmath243 and for observation angles of interest @xmath608
, we have similar reasonings can be done for frequencies around higher harmonics with a different definition of the detuning parameter @xmath599 . however , around odd harmonics , the before - mentioned expansion , together with the application of the resonance approximation for @xmath321 ( fast oscillating terms in powers of @xmath616 $ ] effectively average to zero ) , yields extra - simplifications .
here we are dealing specifically with the first harmonic .
therefore , these extra - simplifications apply .
we neglect both the term in @xmath617 in the phase of eq .
( [ undurad2_ap1 ] ) and the term in @xmath97 in eq .
( [ undurad2_ap1 ] ) .
first , non - negligible terms in the expansion of @xmath612 are those for small values of @xmath618 , since @xmath619 , with @xmath620 .
the value @xmath621 gives a non - negligible contribution @xmath622 .
then , since the integration in @xmath623 is performed over a large number of undulator periods @xmath624 , all terms of the expansion in eq .
( [ undurad3_ap1 ] ) but those for @xmath625 and @xmath626 average to zero due to resonance approximation .
note that surviving contributions are proportional to @xmath627 , and can be traced back to the term in @xmath132 only , while the term in @xmath97 in eq .
( [ undurad3_ap1 ] ) averages to zero for @xmath621 .
values @xmath628 already give negligible contributions .
in fact , @xmath629 . then , the term in @xmath132 in eq .
( [ undurad3_ap1 ] ) is @xmath630 times the term with @xmath621 and is immediately negligible , regardless of the values of @xmath631 .
the term in @xmath97 would survive averaging when @xmath632 and when @xmath633 .
however , it scales as @xmath634 .
now , using condition ( [ resext_ap1 ] ) we see that , for observation angles of interest @xmath608 , @xmath635 . therefore , the term in @xmath97 is negligible with respect to the term in @xmath132 for @xmath621 , that scales as @xmath627 .
all terms corresponding to larger values of @xmath636 are negligible . | we consider the maximum of the wigner distribution ( wd ) of synchrotron radiation ( sr ) fields as a possible definition of sr source brightness
. such figure of merit was originally introduced in the sr community by kim .
the brightness defined in this way is always positive and , in the geometrical optics limit , can be interpreted as maximum density of photon flux in phase space . for undulator and bending magnet radiation from a single electron , the wd function can be explicitly calculated . in the case of an electron beam with a finite emittance
the brightness is given by the maximum of the convolution of a single electron wd function and the probability distribution of the electrons in phase space .
in the particular case when both electron beam size and electron beam divergence dominate over the diffraction size and the diffraction angle , one can use a geometrical optics approach . however , there are intermediate regimes when only the electron beam size or the electron beam divergence dominate . in this asymptotic cases
the geometrical optics approach is still applicable , and the brightness definition used here yields back once more the maximum photon flux density in phase space . in these intermediate regimes
we find a significant numerical disagreement between exact calculations and the approximation for undulator brightness currently used in literature .
we extend the wd formalism to a satisfactory theory for the brightness of a bending magnet .
we find that in the intermediate regimes the usually accepted approximation for bending magnet brightness turns out to be inconsistent even parametrically .
* deutsches elektronen - synchrotron * * + * desy 14 - 129 july 2014 @xmath0 @xmath1 * brightness of synchrotron radiation from undulators and bending magnets * @xmath2 gianluca geloni , _ + european xfel gmbh , hamburg _ vitali kocharyan and evgeni saldin _ + deutsches elektronen - synchrotron desy , hamburg _
@xmath1 issn 0418 - 9833 @xmath1 * notkestrasse 85 - 22607 hamburg * |
the hypothesis that the planets in the solar system were formed in a flattened differentially rotating gaseous disc was originally proposed by laplace ( 1796 ) .
the suggestion arises naturally in view of their orbital configuration .
they are in near circular orbits which approximately lie in the same plane .
models in which planet formation occurs in gaseous discs have been the subject of much theoretical interest in recent times ( e.g. lin & papaloizou 1985 , 1993 and references therein ) . when discussing the large scale physical properties of such a disc , it is convenient to make the idealization that it is infinitesimally thin or completely flat .
when , as in the present solar system , most of the mass lies in a central star of mass @xmath0 the local gravitational acceleration can be taken to be @xmath1 a gaseous element at radius @xmath2 then rotates in circular orbit with angular velocity @xmath3 which , to a good approximation , is obtained from centrifugal balance by equating @xmath4 this gives kepler s law @xmath5 in this approximation , in which forces other than that due to the gravity of the central star are neglected , material rotates conserving its specific angular momentum @xmath6 this then changes only slowly when viscous or other perturbative forces are taken account of .
between 25 and 75 percent of young stellar objects in the orion nebula appear to have discs ( mccaughrean & stauffer 1994 ) .
masses of @xmath7 m@xmath8 and dimensions @xmath9 au have been estimated ( beckwith & sargent 1996 ) .
the presence of discs on the scale of astronomical units has been inferred from the infrared excesses observed in about half of all t tauri stars .
their colors and magnitudes can be fitted by those expected for young pre
main sequence stars with ages @xmath10 yr ( strom et al . 1993 ) . the infrared emission may be produced by the gravitational potential energy liberated by matter flowing inwards at a rate @xmath11 m@xmath12 yr@xmath13 the non observation of discs around older t tauri stars together with these values of @xmath14 suggest a disc lifetime of @xmath15 yr .
it is believed that protostars and protoplanetary discs derive from interstellar matter contained in molecular clouds .
observations ( goodman et al . 1993 ) indicate that typical star forming dense cores in dark molecular clouds have specific angular momentum @xmath16 @xmath17 s@xmath18 . when these clouds undergo gravitational collapse , @xmath19 is initially approximately conserved because the collapse is dynamical .
gas in the outer parts will not be able to fall directly to the centre ( @xmath20 ) if @xmath19 is not zero . to obtain an estimate of the initial size of the disc
, we consider the idealized situation when the pre
collapse cloud is a cold rotating sphere of mass @xmath21 and radius @xmath22 with a single axis of rotation .
matter located on the rotation axis has no angular momentum so it can fall directly to the centre .
an estimate of the time required , @xmath23 is given by the time required to free fall from rest through a distance @xmath24 under the initial inward gravitational acceleration at the surface , @xmath25 this gives @xmath26 in terms of the initial mean density @xmath27 of the cloud , @xmath28 for a cloud core with @xmath29 m@xmath30 and @xmath31 pc , this gives @xmath32 yr .
while matter located on the rotation axis can move directly to the centre , matter in the equatorial plane at the surface of the sphere will be at the outermost radius @xmath33 of a disc after collapse .
conservation of specific angular momentum determines @xmath33 . assuming the total mass , @xmath34 to be concentrated at the centre , the angular velocity at the outer edge of the disc will be given by kepler s law such that @xmath35 thus @xmath33 is given in terms of the conserved quantities @xmath19 and
@xmath21 by @xmath36 adopting @xmath37 @xmath17 s@xmath18 and @xmath38 m@xmath8 we find that @xmath39 au .
the characteristic dimension of such a disc is about an order of magnitude larger than our present solar system but is similar to those of protostellar discs now being observed by direct imaging .
the formation of a protostellar disc through the collapse of a molecular cloud core takes @xmath40@xmath41 yr . during the early stages when the disc is still embedded ( class 0/1 object ) and has a significant mass compared to the central star , there may exist strong disc winds and bipolar outflows ( e.g. reipurth et al .
1997 ) with associated magnetic fields . during this stage
a hydromagnetic disc wind may be an important means of angular momentum removal for the system ( see papaloizou & lin 1995 , and references therein ) .
when the mass of the disc is significant compared to that of the star , there may be a short period ( @xmath42 yr ) of non axisymmetric global gravitational instability with associated outward angular momentum transport ( papaloizou & savonije 1991 ; heemskerk et al .
1992 ; laughlin & bodenheimer 1994 ; pickett et al . 1998 ) that results in additional mass growth of the central star .
this redistribution may occur on the dynamical timescale ( a few orbits ) of the outer part of the disc and so may be quite rapid , on the order of @xmath40 yr for @xmath43 au . the parameter governing the importance of disc self
gravity is the toomre parameter , @xmath44 with @xmath45 being the disc mass contained within radius @xmath2 and @xmath46 being the disc semi thickness . in this review
we shall take @xmath46 to be the distance between the disc mid plane and surface .
this is usually a factor of 23 larger than the vertical extent reached by a disc particle moving through the disc mid
plane with the local sound speed @xmath47 thus @xmath48 typically @xmath49 ( stapelfeldt et al .
1998 ) such that the condition for the importance of self gravity , @xmath50 gives @xmath51 the characteristic scale associated with growing density perturbations in a disc undergoing gravitational instability with @xmath52 is @xmath53 and the corresponding mass scale is @xmath54 which is @xmath55 m@xmath56 for @xmath57 and @xmath58 m@xmath8 with m@xmath56 being jupiter s mass .
gravitational instability does not necessarily lead to fragmentation ( e.g. pickett et al . 1998 ) , nonetheless it has been proposed as a mechanism for directly forming giant planets by cameron ( 1978 ) and more recently by boss ( 1998 ) . during the period of global gravitational instability , it is reasonable to suppose that the disc mass is quickly redistributed and reduced such that global gravitational stability is restored ( @xmath59 ) , after which further disc evolution occurs on a longer timescale governed by viscosity with effects due to self
gravity being small . during this phase ,
the disc may attain a configuration similar to that expected for the minimum mass primordial solar nebula , @xmath60@xmath61 m@xmath62 planets have been proposed to form out of such a disc by a process of growth through planetesimal accumulation followed , in the giant planet case , by gas accretion ( safronov 1969 ; wetherill & stewart 1989 ) . during this evolutionary phase ,
it is reasonable to regard the disc as an axisymmetric configuration in which , to a first approximation , material orbits in circlar keplerian orbits .
however , other weaker forces due to internal pressure , viscosity , or magnetic fields may also operate in the disc . these can result in angular momentum redistribution on a long timescale . in order to flow inwards , material has to transport its angular momentum outwards to matter at larger radii .
the angular momentum transport process determines the timescale on which mass accretion can occur and hence the evolutionary timescale of the disc . historically , the first angular momentum transport mechanism to be considered was through the action of viscosity ( von weizscker 1948 ) .
this acts through the friction of neighbouring sections of the disc upon each other .
the inner regions rotate faster than the outer regions and thus viscous friction tends to communicate angular momentum from the inner parts of the disc to the outer parts . in order to result in evolution on astronomically interesting timescales ,
it is necessary to suppose that an anomalously large viscosity is produced through the action of some sort of turbulence .
the magnitude of the viscosity is usually parameterized through the shakura & sunyaev ( 1973 ) @xmath63 prescription . in this
we suppose the kinematic viscosity coefficient @xmath64 where @xmath63 is a dimensionless constant which must be less than unity .
currently the most likely mechanism for producing turbulence is through hydromagnetic instabilities ( balbus & hawley 1991 ) which might produce @xmath65 provided the disc has adequate ionization . gammie ( 1996 ) has proposed a model in which viscosity only operates in the surface layers where external sources of ionization such as cosmic rays can penetrate .
such layered models may have an interior dead zone of material for @xmath66 au and , although they can be considered as models with a variable @xmath67 they will behave somewhat differently from the standard models considered here which have @xmath63 assumed to be constant throughout .
the evolution of a viscous disc is controlled by angular momentum conservation .
the conservation of specific angular momentum may be expressed in the form ( eg .
papaloizou & lin 1995 ) @xmath68 where @xmath69 is the angular momentum flux , with the vertical coordinate being denoted by @xmath70 , and @xmath71 is the mass density .
the process of vertical averaging applied to equation ( [ cons ] ) in the case of a keplerian disc yields @xmath72 where @xmath73 is the radial velocity in the disc and @xmath74 is the surface mass density . the vertical average for a quantity @xmath75
is defined by @xmath76 we assume @xmath77 vanishes on the disc boundaries ensuring that the total angular momentum of the disc is conserved .
we comment that this implies that angular momentum loss through winds or gain through mass infall is neglected .
the radial angular momentum flux arising from viscosity is given by @xmath78 and thus @xmath79 for a keplerian disc the angular velocity decreases outwards so that the angular momentum flux is directed outwards as required in order that mass accretion may occur .
using equation ( [ flu ] ) in equation ( [ ave ] ) and solving for @xmath80 gives @xmath81 for a keplerian disc in which @xmath82 we get @xmath83 thus the radial velocity in the disc is determined in terms of the kinematic viscosity . for a steady state disc
in which @xmath84 is constant , we find @xmath85 this velocity is negative , consistent with accretion onto the central object . to obtain a general equation governing the disc , the velocity ( [ vrad ] ) is used together with the vertically averaged continuity equation @xmath86 the global evolution of the disc is thus found to be governed by a single diffusion equation for the surface density which takes the form ( lynden bell & pringle 1974 ) @xmath87 according to this the characteristic evolution timescale for the disc will be the global diffusion timescale .
the diffusion coefficient is @xmath88 and , adopting a characteristic radius @xmath89 the global diffusion timescale is @xmath90 for viscosity coefficient @xmath91 parameterized through the @xmath63 prescription so that @xmath92 we obtain @xmath93 for a protostellar disc of size @xmath94 au with @xmath95 and a central solar mass , we find @xmath96 yr .
this gives lifetimes comparable to those estimated for discs around t tauri stars if @xmath97@xmath98 equation ( [ aa4 ] ) enables the evolution of the disc to be determined if @xmath99 is specified as a function of @xmath100 this can be done by solving for the vertical structure of the disc .
we adopt the equation of vertical hydrostatic equilibrium : @xmath101 and the energy equation , which states that the rate of energy removal by radiation is locally balanced by the rate of energy production by viscous dissipation : @xmath102 where @xmath103 is the radiative flux of energy through a surface of constant @xmath70 which is given by : @xmath104 here @xmath105 is the pressure , @xmath106 is the temperature , @xmath107 is the opacity , which in general depends on both @xmath71 and @xmath106 , and @xmath108 is the stefan boltzmann constant . to close the system of equations , we relate @xmath105 , @xmath71 and @xmath106 through the equation of state of an ideal gas :
@xmath109 where @xmath110 is the boltzmann constant , @xmath111 is the mean molecular weight and @xmath112 is the mass of the hydrogen atom . since the main component of protostellar disks at the temperatures we consider is molecular hydrogen , we take @xmath113 . we denote the isothermal sound speed by @xmath114 ( @xmath115 ) . as above
, we adopt the @xmath63parametrization of shakura & sunyaev ( 1973 ) , so that the kinematic viscosity is written @xmath116 . in general
, @xmath63 is a function of both @xmath2 and @xmath70 .
however , we shall limit our calculations presented below to cases with constant @xmath63 . with this formalism ,
equation ( [ dfdz1 ] ) becomes : @xmath117 we have to solve three first order ordinary differential equations for the three variables @xmath103 , @xmath105 ( or equivalently @xmath71 ) , and @xmath106 as a function of @xmath70 at a given radius @xmath118 accordingly , we need three boundary conditions at each @xmath2 .
we shall denote with a subscript @xmath119 values at the disk surface .
the first boundary condition is obtained by integrating equation ( [ dfdz1 ] ) over @xmath70 between @xmath120 and @xmath46 . since by symmetry
@xmath121 , this gives : @xmath122 where we have defined @xmath123 if the disk were in a steady state , @xmath124 would not vary with @xmath2 and would be the constant accretion rate through the disk . in general however , this quantity does depend on @xmath2 .
the second boundary condition is determined by the fact that very close to the surface of the disc , since the optical depth @xmath125 above the disc is small , we have : @xmath126 this condition is familiar in stellar structure ( e.g. schwarzschild 1958 ) . using @xmath127 we thus obtain @xmath128 provided @xmath129 the results do not depend on the value of @xmath125 we choose ( see below ) .
the third boundary condition we use is @xmath130 here the disc is assumed immersed in a medium with background temperature @xmath131 the surface opacity @xmath132 in general depends on @xmath133 and @xmath134 and we have used @xmath135 .
the boundary condition ( [ ts ] ) is the same as that used by levermore & pomraning ( 1981 ) in the eddington approximation ( their eq .
[ 56 ] with @xmath136 ) . in the simple case when @xmath137 and the surface dissipation term involving @xmath63 is set to zero
, it simply relates the disc surface temperature to the emergent radiation flux . at a given radius @xmath2 and for a given values of the parameters @xmath124 and @xmath63 , we solve equations ( [ dpdz ] ) , ( [ dtdz ] ) and ( [ dfdz ] ) with the boundary conditions ( [ fs ] ) , ( [ ps ] ) and ( [ ts ] ) to find the dependence of the state variables on @xmath138 the opacity is taken from bell & lin ( 1994 ) .
this has contributions from dust grains , molecules , atoms and ions .
it is written in the form @xmath139 where @xmath140 , @xmath141 and @xmath142 vary with temperature .
the equations are integrated using a fifth order runge
kutta method with adaptive step length ( press et al .
1992 ) . for a specified @xmath143
we determine the value of @xmath144 the vertical height of the disc surface .
this is done iteratively .
starting from an estimated value of @xmath144 after satisfying the surface boundary conditions , the equations are integrated down to the mid plane @xmath145 the condition that @xmath146 at @xmath147 will not in general be satisfied .
an iteration procedure ( e.g. the newton raphson method ) is used to adjust value of @xmath46 until @xmath148 at @xmath147 to a specified accuracy . an important point to note
is that as well as finding the disc structure , we also determine the surface density @xmath149 for a given @xmath150 thus a @xmath151 relation is derived . in the calculations presented here ,
we have taken the optical depth of the atmosphere above the disk surface @xmath152 and a background temperature @xmath153 k. in these calculations the temperatures are lower than about 4,000 k , so that , at the densities we consider , hydrogen is not dissociated and the mean molecular weight @xmath113 . in the optically thick regions of the disk ,
the value of @xmath46 is independent of the value of @xmath125 we choose .
however , this is not the case in optically thin regions where we find that , as expected , the smaller @xmath154 the larger @xmath155 however , this dependence of @xmath46 on @xmath125 has no physical significance , since in all cases , the mass is concentrated towards the disk mid plane in a layer with thickness independent of @xmath156 for example , for @xmath157 m@xmath30 yr@xmath18 and @xmath158 , we find , at @xmath159 au , that @xmath160 and 0.24 for @xmath152 and @xmath161 , respectively .
however , in both cases , the surface density , the optical thickness and the mid plane temperature are the same . only the mid plane pressure varies slightly ( by about 30% ) between these cases . in figures [ fig1]a c , we plot @xmath162 , @xmath149 and the mid plane temperature @xmath163 versus @xmath2 for @xmath124 between @xmath164 and @xmath165 m@xmath30/year ( assuming this quantity is the same at all radii , i.e. the disk is in a steady state ) and for illustrative purposes we have adopted @xmath166 .
figure [ fig1]a indicates that the outer parts of the disk are shielded from the radiation of the central star by the inner parts .
this is in agreement with the results of lin & papaloizou ( 1980 ) and bell et al .
( 1997 ) . for @xmath166 ,
the radius beyond which the disk is not illuminated by the central star varies from 0.2 au to about 3 au when @xmath124 goes from @xmath164 to @xmath165 m@xmath30 yr@xmath18 .
these values of the radius move to 0.1 and 2 au when @xmath158 .
since reprocessing of the stellar radiation by the disk is not an important heating factor below these radii , this process will in general not be important in protostellar disks .
we note that this result is independent of the value of @xmath125 we have taken .
indeed , as we pointed out above , only the thickness of the optically thin parts of the disk ( which do not reprocess any radiation ) gets larger when @xmath125 is decreased .
the values of @xmath162 , @xmath149 and @xmath163 we get are qualitatively similar to those obtained by lin & papaloizou ( 1980 ) , who adopted a prescription for viscosity based on convection , and bell et al .
our values of @xmath162 are somewhat larger though , since @xmath46 is measured from the disk mid plane to the surface such that @xmath125 is small , and not 2/3 as usually assumed .
however , as we commented above , this has no effect on the other physical quantities .
we also recall that @xmath144 as defined here , is about 23 times larger than @xmath167 with @xmath114 being the mid plane sound speed , which is often used to define the disc semi thickness .
we determine the evolution of the radial structure of a non steady @xmath63disk by solving the diffusion equation ( [ aa4 ] ) . to do this , we need to use the relation between @xmath168 and @xmath149 at each radius .
interpolation of or piece wise power law fits to numerical data may be used to represent this relation and more details of these will be published elsewhere .
we note that they can be used either to compute @xmath149 from @xmath124 or @xmath124 from @xmath100 in figures [ fig2]a b we plot both the curves @xmath169 that we get from the vertical structure integrations as described above and those obtained from piece wise power law fits .
figures [ fig2]a and [ fig2]b are for @xmath158 and @xmath170 , respectively . in each case
the radius varies between 0.01 and 100 au .
if we calculate @xmath124 using the fits with @xmath149 as an input parameter , the average error is 22 , 18 , 13% and the maximum error is 55 , 48 , 42% for @xmath171 and @xmath172 , respectively .
we see that the fits give a good approximation . using the @xmath173 relation derived from the integrations , we solved equation ( [ aa4 ] ) using explicit finite difference techniques .
we considered the situation of a disc with initially @xmath174 m@xmath30 for which @xmath175 extending to 100 au .
the central star had @xmath176 m@xmath30 and for illustrative purposes we adopted @xmath177 in figure [ fig3 ] we show the evolution of @xmath149 as a function of time .
after a time @xmath178 yr , the @xmath149 profile resembles that of a steady disc in the inner parts justifying the assumption of a steady state disc model . after this time
, the model is similar to that assumed for the primordial solar nebula with @xmath179 g @xmath180 at 5 au .
it is thought that planetesimals can be built up from @xmath181 sized particles through processes of collision , sticking and accumulation occurring in a gaseous medium with some degree of turbulence ( see the review by weidenschilling & cuzzi 1993 ) .
if physical parameters are favourable , particles with a size distribution ranging up to @xmath182 a few km can be produced on timescales on the order of @xmath183 yr at 1au .
the efficiency of these processes are very uncertain depending as they do on sticking probabilities and the degree of particle settling in a turbulent medium etc .
planetesimal formation efficiency may be a function of the nebula location , being more effective beyond a few au , where ice has condensed .
here we assume that planetesimals with mass @xmath184 g may form on a sufficiently rapid timescale anywhere in the nebula .
we suppose the planetesimals have number distribution @xmath185 for some index @xmath186 here the number of planetesimals in the mass range @xmath187 is @xmath188 then most of the mass is distributed in the most / least massive objects according as @xmath189 or @xmath190 the surface density of matter in the form of planetesimals is @xmath191 the characteristic number density , @xmath192 for characteristic mass @xmath193 is then @xmath194 here @xmath195 is the scale height of the distribution @xmath196 where @xmath197 is the root mean square velocity dispersion . for the typical values , @xmath198 g @xmath199 @xmath200 g , and @xmath201 cm ,
the mean distance between planetesimals is @xmath202 cm .
this suggests use of a local box model ( e.g. stewart & wetherill 1988 ) . in this box
, the centre of which is in circular orbit with the local keplerian angular velocity , planetesimals are assumed to move between encounters with constant velocity . in this respect , the effect of the central mass is ignored , and the planetesimals are treated using the methods of kinetic theory . the idea ( safronov 1966 ) is that the velocity dispersion of the planetesimals is increased by gravitational scattering , enhancing direct collisions between them through which they accumulate and grow .
the local box model fails when the largest planetesimals become isolated .
that happens when accumulation has reached the stage where there are so few that they are in non overlapping near circular orbits such that they can not perturb similar mass objects in neighbouring orbits into collision . in this situation
the central mass can no longer be neglected . to get an approximate idea of when isolation occurs at radius @xmath89
equate @xmath203 this is a statement that there is one object in an annulus of width equal to four times its roche lobe size .
then @xmath204 this gives @xmath205 g at 5 au for @xmath198 g @xmath180 and @xmath206 m@xmath30 with obvious scalings to other surface densities and radii .
thus on the order of 1 m@xmath207 may be obtained at 5au .
this argument assumes circular orbits which is probably reasonable for the largest objects which are circularized through dynamical friction . after the isolation stage , planetesimal evolution is probably best followed by global n body methods ( e.g. aarseth et al .
1993 ) . in the gas free case
the ultimate result is expected to be the formation of terrestrial planets . for a member of a swarm of equal masses interacting gravitationally ,
the cross section for elastic scattering is on average larger than that for a direct collision , so long as the velocity dispersion is less than the escape velocity .
thus when the velocity dispersion of the swarm is initially small , it will increase because of the effect of elastic scatterings . the interaction radius for two masses @xmath193 to give significant scattering is @xmath208 the time between encounters is then @xmath209 here the @xmath210 term accounts for more distant collisions ( binney & tremaine 1987 ) . using the above , one gets @xmath211 for @xmath212 g , @xmath198 g @xmath180 and @xmath206 m@xmath30 , one finds at 1 au that @xmath213 yr .
note that @xmath195 is small because the planetesimal dispersion velocities are expected to reach the escape velocity for the characteristic mass at a maximum .
at this point inelastic physical impacts become as important as scattering and damp the random motions . then for @xmath212 g , @xmath214 cm s@xmath18 , @xmath215 and @xmath216 yr at 1 au .
thus early planetesimal build up occurs on a rapid timescale . from the above arguments , the collision time of @xmath193 with @xmath217 is inversely proportional to @xmath218
if this does not decrease with @xmath217 , then collisions with larger masses are dominant and we expect velocity dispersions to build up to the escape velocity of the largest body ( safronov 1966 ) .
this occurs for @xmath219 for @xmath220 the largest bodies collide with predominantly smaller ones and are circularized by dynamical friction .
they can then accrete efficiently from the smaller ones which move with a velocity dispersion small compared to the escape velocity from them .
the accretion rate from planetesimals with mass @xmath217 and velocity dispersion @xmath221 is enhanced by gravitational focusing .
thus the growth of the largest mass @xmath193 say , with radius @xmath222 is given by @xmath223 runaway is caused both by the increase of @xmath222 with @xmath193 and the gravitational focusing term in brackets .
the timescale for growth to isolation mass is comparable to the encounter time , @xmath224 indicated above .
but note that growth may be slowed down through the effect of encounters between neighbouring runaways producing an increase in the velocity dispersion that propagates to all components of the system ( ida & makino 1993 )
after a solid protoplanetary core grows to a critical mass of around a few @xmath225 the surrounding gaseous atmosphere can no longer grow quasi statically in mass along with it . a process of collapse ensues possibly leading to dynamical accretion and mass growth to values characteristic of giant planets .
we review the theory of this below .
such a critical core mass model is supported by the indication from models of jupiter that it has a solid core of @xmath226@xmath227 m@xmath207 ( podolak et al .
1993 ) .
let @xmath228 be the spherical polar radius in a frame with origin at the centre of the planet s core .
we neglect the rotation of the planet around both its own spin axis and the disc spin axis .
we assume that the envelope is in hydrostatic equilibrium and spherically symmetric , so that : @xmath229 where @xmath230 is the acceleration due to gravity , @xmath231 being the mass contained in the sphere of radius @xmath228 ( this includes the core mass if @xmath228 is larger than the core radius )
. mass conservation gives : @xmath232 the thermodynamic variables in the protoplanet envelope are such that the equation of state of an ideal gas does not normally apply .
here we adopt the state of the art chabrier et al .
( 1992 ) equation of state for a hydrogen and helium mixture .
we fix the abundances of hydrogen and helium to be 0.7 and 0.28 respectively .
the equation of radiative transport is : @xmath233 where @xmath234 is the luminosity carried by radiation . denoting the radiative and adiabatic temperature gradients by @xmath235 and @xmath236 respectively , we have @xmath237 and @xmath238 with the subscript @xmath119 meaning that the derivative has to be evaluated at constant entropy .
we assume that the only energy source comes from the core which outputs the core luminosity @xmath239 , given by : @xmath240 where @xmath241 and @xmath242 are respectively the mass and the radius of the core , and @xmath243 is the rate of accretion of planetesimals onto the core .
we note that it is customary to take , instead of @xmath239 , the luminosity supplied by the gravitational energy which the planetesimals entering the planet atmosphere release near the surface of the core ( see , e.g. , mizuno 1980 ; bodenheimer & pollack 1986 )
. however , when the mass of the atmosphere is small compared to that of the core , these two luminosities are comparable , providing we take for @xmath243 the rate of accretion of planetesimals onto the core during the phase of accretion of the atmosphere and not during the phase of the core formation . these two rates may be different if the core has migrated in the disk before accreting the atmosphere .
if @xmath244 , the medium is convectively stable and the energy is transported only by radiation . in that case @xmath245 .
when @xmath246 , some energy is transported by convection . in that case ,
@xmath247 , where @xmath248 is the convective luminosity .
we use the expression for @xmath248 given by mixing length theory ( cox & giuli 1968 ) : @xmath249^{3/2 } \sqrt { \frac{1}{2 } \rho g \left| \left ( \frac{\partial \rho}{\partial t } \right)_p \right| } , \ ] ] where @xmath250 is the mixing length , @xmath251 being a constant of order unity , @xmath252 , and the subscript @xmath105 means that the derivative has to be evaluated for a constant pressure .
the quantities @xmath253 and @xmath236 are given by chabrier et al .
( 1992 ) .
we suppose that the planet core has a uniform mass density @xmath254 here taken to be 3.2 g @xmath255 the core radius , which is the inner boundary of the atmosphere , is then given by : @xmath256 the outer boundary of the atmosphere is taken to be at the roche lobe radius @xmath257 of the planet : @xmath258 where @xmath259 is the planet mass , @xmath260 being the mass of the atmosphere , and @xmath2 is the location of the planet in the disk ( i.e. the separation between the planet and the central star ) . to avoid confusion , we will denote the disk mid plane temperature , pressure and mass density at the distance @xmath2 from the central star by @xmath261 @xmath262 and @xmath263 , respectively . at @xmath264 the mass is equal to @xmath265 , the pressure is equal to @xmath262 and the temperature is given by : @xmath266 where
@xmath267 the condition at @xmath268 is that the mass is equal to @xmath241 there . at a given disk radius @xmath2 and for a given core mass @xmath269
we solve the equations ( [ dpdvarpi ] ) , ( [ dmdvarpi ] ) and ( [ dtdvarpi ] ) with the boundary conditions described above to get the structure of the envelope .
the opacity law adopted was the same as that for the disk models .
we note that when the density gets large , the interior of the envelope becomes convective so that the value of the opacity does not matter there .
the equations are integrated using the fifth order runge
kutta method with adaptive step size control ( press et al .
we guess a starting value of @xmath260 and integrate the equations from @xmath270 down to the core surface @xmath268 .
we then iterate the integration , adjusting @xmath260 at each step , until the solution gives @xmath271 at @xmath268 with some accuracy . at each radius @xmath89 for a fixed @xmath243 , there is a critical core mass @xmath272 ( which increases as @xmath243 increases ) above which no solution can be found , i.e. there can be no atmosphere in hydrostatic and thermal equilibrium confined between the radii @xmath242 and @xmath257 around cores with mass larger than @xmath272 .
this is because when the core mass is too large , the atmosphere has to collapse onto the core in order to supply adequate luminosity to support itself . for masses below @xmath272 , there are ( at least ) two solutions , corresponding to a low mass and a high mass envelope respectively
. in figure [ fig4 ] we plot @xmath265 versus @xmath241 for different @xmath243 ( between @xmath165 and @xmath273 m@xmath207 yr@xmath18 ) at a radius of 5 au and for @xmath274 k and @xmath275 dyn @xmath180 .
these values of the temperature and pressure are obtained from the vertical structure integrations described above when the parameters @xmath158 and @xmath276 m@xmath30 yr@xmath18 are used at @xmath277 au . the critical core mass , which decreases when @xmath243 decreases , is found to be 22.5 m@xmath207 for @xmath278 m@xmath207 yr@xmath18 and 2.5 m@xmath207 for @xmath279 m@xmath207 yr@xmath18 .
these values are slightly larger than those found by bodenheimer & pollack ( 1986 )
. the difference may be accounted for by the fact that we do not calculate the luminosity in exactly the same way as they do .
also we use a slightly different boundary condition for the temperature at the surface of the planet .
once the planetary mass has attained values of around an earth mass or higher , dynamical interactions with the surrounding disc matter become important , leading to phenomena such as inward orbital migration and gap formation ( lin & papaloizou 1993 ; ward 1997 ; lin et al .
1998 ) .
korycansky & papaloizou ( 1996 ) considered the perturbed disc flow around an embedded protoplanet when the imposed viscosity @xmath280 they used a shearing sheet approximation in which a patch , centered on the planet , corotating with its orbit is considered in a 2d approximation .
for unit of length @xmath281 was adopted , where @xmath2 is the planet s orbital radius .
we remark that @xmath282 is a multiple of the roche lobe radius used above .
when the basic equations are expressed in dimensionless units , the only parameter defining the problem is @xmath283 being essentially the ratio of roche lobe radius to disc semi thickness .
the velocity @xmath284 viewed in the rotating frame was split into components involving a vortical function @xmath285 and a potential @xmath286 : @xmath287 for a steady flow , there is also a stream function @xmath288 such that @xmath289 parameters relating to the flow for @xmath290 are plotted in figures 5 and 6 . here
cartesian coordinates are adopted with origin at the centre of the protoplanet and @xmath291 axis pointing along the line joining the central star to the protoplanet . the perturbed surface density ( figure 5 ) shows trailing shock waves behind which there is a strong surface density enhancement giving rise to pronounced wakes . also notable
is the prograde disc flow around the protoplanet ( figure 6 ) .
dissipation in the shock waves , which become strong once @xmath292 eventually results in gap formation ( lin & papaloizou 1993 ) .
protoplanet disc interaction leads to gap formation and orbital migration , which , together with tidal interaction with the central star ( terquem et al .
1998 ) can lead to massive planets in circular orbits , with periods of a few days , as observed ( butler et al .
the reader is referred to the article by lin et al .
( 1998 ) for an account of the dynamical phenomena which can play a role in determining the final orbital configuration of planetary systems . | the formation , structure and evolution of protoplanetary discs is considered .
the formation of giant planets within the environment of these models is also discussed . |
the combined statistical analysis of the most recent measurements coming from jla ( joint light curve analysis ) type ia supernovae ( sne ia ) data @xcite , the growth rate of structure formation obtained from redshift space distortion ( rsd ) data @xcite , the different baryon acoustic oscillation ( bao ) detected in the galaxy clustering observations ( @xmath8dfgs , sdss dr @xmath9 , sdss dr @xmath10 , sdss dr @xmath11 , boss dr @xmath10 cmass , @xmath12dspcf , @xmath12dmps , boss dr @xmath11 cmass ) , @xcite , the observations of anisotropies in the power spectrum of the cosmic microwave background ( cmb : distance priors ) data from the planck @xmath13 data @xcite , and the hubble parameter ( h ) measurements obtained from galaxy surveys @xcite indicate that the present universe is undergoing a phase of accelerated expansion . from the theoretical point of view
, this phenomenon can be explained introducing in the universe an unknown physical fluid with negative pressure so - called @xmath2 @xcite .
many alternative models have been suggested to explain it ; in particular , the lambda cold dark matter ( @xmath14cdm ) model has a cosmological constant as @xmath2 with an eos parameter @xmath15 @xcite .
however , other more structured models replace @xmath14 by a dynamical @xmath2 such as phantom model @xcite , quinton model @xcite , quintessence model @xcite , k - essence model @xcite , chaplyging gas model @xcite , massive scalar field model @xcite , etc .
+ in the same way , within the universe we also assume the existence of another dark component so - called @xmath1 , which acts exactly like the ordinary matter ( pressureless ) , and can interact with @xmath2 gravitationally .
+ on the other hand , the @xmath14cdm model presents two different problems such as the fine tuning and the cosmic coincidence .
then , one way to solve the last problem within gr is to assume a coupling between @xmath2 fluid and @xmath1 fluid . currently , there are not neither physical arguments nor recent observations to exclude an energy exchange between these dark components because their natures are still unknown @xcite .
+ due to the absence of a fundamental theory to construct @xmath0 , different phenomenological parameterizations for @xmath0 have been proposed by mathematical simplicity @xcite ; for example , @xmath16 @xcite , @xmath17 @xcite , @xmath18 @xcite , @xmath19 @xcite and @xmath20 @xcite .
then such models may be physically viable , if they are confronted with the observational data , and therefore , can be employed in order to look new physical properties on cosmological scales @xcite .
+ on the other hand , the properties of @xmath2 fluid are mainly characterized by the eos parameter @xmath3 . in this case , two possibilities are proposed to explain a varying @xmath3 . the first one is to parameterize @xmath3 in terms of some free - parameters @xcite .
thus , among all the different ansatzes the chevallier - polarski - linder ( cpl ) parameterization @xcite @xmath21 $ ] ( where @xmath22 is the redshift , @xmath23 and @xmath24 are dimensionless parameters ) is considered as the most popular ansatz .
this ansatz shows a divergence problem when redshift @xmath22 approaches to @xmath25 @xcite .
the second one is to expand @xmath3 in terms of an appropriated local basis @xcite .
consequently , we are interested in proposing a divergence - free reconstruction for @xmath3 and via a polynomial expansion , it will show new features .
for example , we can expand @xmath3 , in function of the chebyshev polynomials @xmath26 , with @xmath27 .
these polynomials are considered as a complete orthonormal basis on the finite interval @xmath28 $ ] , and belong to the hilbert space @xmath29 of real values @xcite .
likewise , they have the property to be minimal approximately polynomials @xcite , and possess a better advantage in terms of stability .
+ if we compare the theoretical predictions with the observational measurements , we will show the different effects of including the numerical reconstructions of @xmath0 and @xmath3 on the energy densities ( @xmath30 and @xmath31 ) , on the evolution of the linear growth rate of @xmath1 density perturbation ( @xmath4 ) , on the effective hubble friction term ( @xmath5 ) and on the effective newton constant ( @xmath6 ) , respectively .
this will be the main aim of the present work .
+ on the other hand , rsd data represent a compilation of measurements of the quantity @xmath32 at different redshifts , which were obtained in a model independent way .
these data are apparent anisotropies ( effects ) of the galaxy distribution ( in redshift space ) , due to the differences of the estimates between the redshifts observed distances and true distances , and caused by the component along the line of sight ( los ) of the peculiar velocity of each of the galaxies ( recessional velocity ) @xcite .
therefore , rsd data will provide tight constraints on the parameter space of the cosmic scenarios , and the necessary information to discriminate among all them @xcite .
+ furthermore , another interesting observable considered here are the measurements of the alcock - paczynski ( ap ) effect @xcite .
this ap test describes a distortion along the observed tangential and radial dimensions of objects , which are assumed as isotropic @xcite .
this signal depends on the value of the @xmath33 parameter , and will be very useful to constrain cosmological models . in this work ,
the linear growth rate @xmath4 is constrained by measuring the rsd signal , while the dilation scale ( @xmath34 ) @xcite and @xmath33 parameter @xcite evaluated at an effective redshift @xmath35 are constrained by measuring the bao and ap signals , respectively .
+ now then , two distinct coupled @xmath2 models such as xcpl and dr are studied here , and from which we have found a determined number of different effects such as a reduction or enhancement on the amplitudes of @xmath30 , @xmath31 , @xmath4 , @xmath5 and @xmath6 at large and small redshifts with respect to those found in the uncoupled models
. however , these modifications should be small in order to do not have a significant impact on the matter density perturbations .
furthermore , important features of the universe can be obtained from these changes , and therefore these variations depend of the chosen parameterizations for @xmath0 and @xmath3 , respectively . in this article ,
all our models are constrained using an analysis combined of jla ( sne ia ) @xcite , the growth rate of structure formation obtained from rsd data @xcite , bao data @xcite , cmb data @xcite and the h data set @xcite .
+ finally , we organize this paper as follows : the background equation of motions for the energy densities are presented in section ii . in section iii
we describe the reconstruction schemes for @xmath0 and @xmath3 , respectively . in section
iv we show the theoretical @xmath2 models . in section v we studied the conditions for the crossing of @xmath36 line , and define the redshift crossing points .
the perturbed equation of motions and the equations of structure formations are described in section vi .
the current observational data and the priors considered are presented in section vii .
we discuss our results in section viii . in section ix , we conclude our main results .
in a flat friedmann - robertson - walker ( frw ) universe its background dynamics is described by the following set of equations for their energy densities ( detailed calculations are found in @xcite , so we do not discuss them here . )
@xmath37 where @xmath38 , @xmath39 , @xmath40 and @xmath41 are the energy densities of the baryon , radiation , @xmath1 and @xmath2 , respectively .
now , defined the hubble expansion rate as @xmath42 , and `` @xmath43 '' indicates differentiation with respect to the cosmic time @xmath44 .
+ in what follows , we shall assume that there is not energy transfer from @xmath2 ( @xmath1 ) to baryon or radiation , and among them only exist a gravitational coupling @xcite .
for convenience , we defined the critical density @xmath45 and the critical density today @xmath46 ( in where @xmath47 is the current value of the hubble parameter ) . considering that @xmath48 , and then the normalized densities become @xmath49 the first friedmann equation is given by @xmath50\;,\end{aligned}\ ] ] and with the following relation for all time @xmath51 the scale factor @xmath52 is related to the redshift through @xmath53 ,
from which find @xmath54 .
substituting this last relation into eqs.([eofb])-([eofde ] ) , and solving eqs.([eofb ] ) and ( [ eofr ] ) , we find @xmath55 these equations are fundamental to determine the results within our models .
the chebyshev polynomials form a complete set of orthonormal functions on the interval @xmath57 $ ] , and have the property to be the minimal approximating polynomials . it is to say
, they have the smallest maximum deviation from the true function at any given order @xcite .
+ because of the unknown of the origin and nature of the dark fluids , it is not possible to derive @xmath0 from fundamental principles , but we have the freedom of choosing any possible form of @xmath0 that satisfies eqs . ( [ edfdmomega ] ) and ( [ edfdeomega ] ) simultaneously .
hence , we propose a phenomenological form for a varying @xmath58 , which could be definitely a function of @xmath40 multiplied both by a quantity with units of inverse of time ( for instance @xmath59 ) and by the coupling term , @xmath60 .
since @xmath60 can be modelled as a varying function of @xmath22 and used to measure the strength of the interaction , it can be reconstructed conveniently in terms of chebyshev polynomials .
accordingly , we can look new physical properties in spite of the fact that @xmath0 may be determined by the universal expansion rate @xmath59 .
therefore , this scenario will simplify the analytic results , if we reconstruct @xmath60 from observational data .
an energy exchange in the dark sector can be chosen as @xmath61 here , the strength of the coupling is characterized by @xmath62 where the coefficients of the polynomial expansion @xmath63 are free dimensionless parameters @xcite , and @xmath64 represent the first three chebyshev polynomials . + within the cpl model , the past evolution history may be successfully described by its eos parameter , @xmath3 , but the future evolution may not be explained , because @xmath3 grows increasingly , and then , encounters a divergence when @xmath65 .
that is not a physical feature .
consequently , to avoid such divergence problem we propose here a complete phenomenological reconstruction of a smoothly varying eos parameter , @xmath3 , which can also be expanded in terms of an expansion of the chebyshev polynomials such that @xmath66 where @xmath67 and @xmath68 are free dimensionless parameters .
the polynomial @xmath69 and the parameter @xmath68 were included conveniently to simplify the calculations . however , this suitable generalization should be compatible with recent observational data .
likewise , @xmath70 behaves nearly linear at low redshift @xmath71 and @xmath72 , whereas in the high redshift regime @xmath73 .
+ the chebyshev polynomials of order @xmath74 were defined by eq .
( [ chebyshev1 ] ) .
thereafter , using numerical simulations we will compute the best fitted values for @xmath75 , @xmath76 , @xmath77 , @xmath23 , @xmath24 and @xmath68 , respectively .
in this scenario , the function @xmath78 is defined fixing both @xmath79 , and @xmath80 into eqs .
( [ omegabs])-([edfdeomega ] ) @xmath81\;.\ ] ] within this model , @xmath78 is determined replacing both @xmath82 $ ] , where @xmath23 , @xmath24 are real parameters , and @xmath80 into eqs .
( [ omegabs])-([edfdeomega ] ) @xmath83\;.\end{aligned}\ ] ] firstly a coupled model can be defined putting both @xmath84 , where @xmath23 , @xmath24 are real free parameters , and @xmath85 given by eqs .
( [ interaction ] ) and ( [ couplingiq ] ) into eqs . ( [ omegabs])-([edfdeomega ] ) . the explicit form for @xmath86 and @xmath87
are reached by solving eqs .
( [ edfdmomega ] ) and ( [ edfdeomega ] ) , respectively .
\label{de_xcpl}\omega^{\star}_{de}(z)=(1+z)^{3(1+{\omega}_{0}+{\omega}_{1})}\biggl[{\omega_{de,0}}{\rm exp}\left(\frac{-3\omega_{1}z}{1+z}\right)+\;\quad\qquad\nonumber\\ \frac{z_{max}}{2}\omega_{dm,0}{\rm exp}\left(\frac{3\omega_{1}}{1+z}\right)\sum_{n=0}^{2}\lambda_{n}s_{n}(z,\bar{\omega})\biggr]\;.\quad\qquad\end{aligned}\ ] ] the following average integrals are also defined @xmath89 where we also defined the following expressions for all @xmath90 $ ] ( see appendix [ integralin ] and @xcite ) @xmath91 with the quantities @xmath92 where @xmath93 is the maximum value for @xmath22 , and in which the observations are possible such that @xmath94 $ ] and @xmath95 , respectively . + therefore , the function @xmath78 can be constructed from eqs .
( [ hubble ] ) , ( [ omegabs ] ) , ( [ omegars ] ) , ( [ dm_xcpl ] ) and ( [ de_xcpl ] ) .
secondly another coupled model can be modeled setting @xmath96 where @xmath23 , @xmath24 and @xmath68 are real parameters .
moreover , @xmath97 is given by eqs .
( [ interaction ] ) and ( [ couplingiq ] ) . the analytic form for @xmath86
can be reached by solving eq .
( [ edfdmomega ] ) . for this model eq .
( [ dm_xcpl ] ) represents the solution of eq .
( [ edfdmomega ] ) . instead , the solution of eq .
( [ edfdeomega ] ) can be obtained by numerical integration , from using eq .
( [ a1 ] ) and appendix [ integralin ] , @xmath98 where the following relations are defined @xmath99^{a_1}\;,\nonumber\\ c_2(z)&=&{\rm exp}\biggl[a_{2}\sqrt{2}\left(\arctan\left[\frac{\sqrt{2}[2z - z_{max}]}{2z_{max}+4}\right]+\arctan\left[\frac{\sqrt2z_{max}}{2z_{max}+4}\right]\right)\biggr]\;,\nonumber\\ d_0(z)&=&0.5z_{max}\omega_{dm,0}(1+z)^{2\omega_{0}-2\omega_{1}+5\omega_{2}+3}(2+z^{2})^{\omega_{1}-\omega_{0}+5\omega_{2}}\;,\nonumber\\ d_1(z)&=&{\rm exp}\biggl[{\sqrt{2}}\biggl(\omega_{0}+2\omega_{1}-5\omega_{2}\biggr)\arctan\left(0.5{\sqrt{2}}z\right)\biggr]\;,\;\;\;d_2(z)={\rm exp}\biggl[\arctan\left(\frac{\sqrt{2}z_{max}}{4 + 2z_{max}}\right)j_{2}\biggr]\;,\nonumber\\ f_0(\tilde{x})&=&\frac{{\rm exp}\biggl[-\sqrt{2}\biggl(\omega_{0}+2\omega_{1}-5\omega_{2}\biggr)\arctan\biggl[0.25\sqrt{2}z_{max}(1+\tilde{x})\biggr]\biggr]}{\biggl[\left(\tilde{x}^{2}+2\right)(0.5z_{max})^{2}+2\left(1+z_{max}\right)\biggr]^{\omega_{1}-\omega_{0}+5\omega_{2}}}\;,\nonumber\\ f_1(\tilde{x})&=&\biggl[\frac{\tilde{x}^{2}+2 + 8[z_{max}^{-2}+z_{max}^{-1}]}{3 + 8[z_{max}^{-2}+z_{max}^{-1}]}\biggr]^{j_{1}}\;,\;\;\;f_2(\tilde{x})={\rm exp}\biggl[\arctan\left(\frac{\sqrt2z_{max}\tilde{x}}{2z_{max}+4}\right)j_{2}\biggr]\;,\nonumber\\ g(\tilde{x})&=&\biggl[1 + 0.5z_{max}(1+\tilde{x})\biggr]^{2(\omega_{0}-\omega_{1}+\omega_{2})-j_{0}}\;.\end{aligned}\ ] ] within this model , the function @xmath78 can be constructed from eqs .
( [ hubble ] ) , ( [ omegabs ] ) , ( [ omegars ] ) , ( [ dm_xcpl ] ) and ( [ de_dr ] ) . the basic analytical expressions for @xmath100 and @xmath101 ( when @xmath102 ) are shown in appendix [ integralin ] .
let us now proceed with the calculation of the redshift crossing points , and analyze the behavior of @xmath60 and its derived . from eqs .
( [ interaction ] ) , ( [ couplingiq ] ) and ( [ chebyshev1 ] ) , we note that exist real values of @xmath22 that lead to @xmath104 @xmath105 in where the @xmath106 denotes the redshift crossing points of @xmath103 line .
+ then , the solution to eq .
( [ crossiq ] ) is given by @xmath107 from here , we note that these results depend of the choice for @xmath60 .
however , we are interested in the case where @xmath108 ; in particular , this happens when @xmath103 .
furthermore , the reverse situation is also possible . in this discussion , to guarantee the possibility of the crossing of @xmath103 line we must explore the function @xmath109 .
now , we consider the possibility to have @xmath103 and @xmath110 could be zero or different of zero . then , from eqs .
( [ couplingiq ] ) and ( [ chebyshev1 ] ) , we obtain @xmath111 substituting eq .
( [ zcrossiq ] ) into eq .
( [ dcrossiqdz ] ) , we get the following @xmath112 now , if we consider the possibility to have @xmath104 and @xmath113 , it means that the impossibility for having @xmath114 over @xmath115 line .
by contrast , the only possibility for a crossing corresponds to @xmath116 from eq .
( [ diqdznull2 ] ) , we impose the following restraint for the avoidance of imaginary values in @xmath117 @xmath118 which can be rewritten as @xmath119 similarly , from eq .
( [ diqdznull4 ] ) to guarantee real values with physical sense , we impose @xmath120 the values of @xmath106 for the coupled @xmath2 models are given in table [ crossing_state ] .
let us make some commentaries about the signs of @xmath0 , @xmath60 and @xmath121 , respectively .
in general , if the parameters @xmath122 , @xmath123 and @xmath124 hold positive or negative values , then @xmath60 and @xmath121 will be ambiguously positive , negative or zero , in any epoch of the universe . from eqs .
( [ dcrossiqdz ] ) , ( [ diqdznull4 ] ) , ( [ diqdznull6 ] ) and ( [ diqdznull7 ] ) , if @xmath122 and @xmath124 are both positive or are both negative , then @xmath121 could be positive or negative .
moreover , @xmath121 may be zero when @xmath122 , @xmath123 , and @xmath124 are all zero ( i.e. uncoupled @xmath2 models ) or when @xmath125 .
here we can describe the signs of @xmath0 and @xmath60 , choosing positive values for @xmath122 either with negative values of @xmath123 and positive values of @xmath124 or with positive values of @xmath123 and negative ( positive ) values of @xmath124 , in determined redshift ranges .
let us consider a spatially flat universe with scalar perturbations about the background . in the absence of the anisotropic stress , the perturbed line element in the newtonian gauge
is given by @xcite @xmath126 where @xmath127 is the gravitational potential , @xmath128 is the scale factor , @xmath129 is the background four velocity , @xmath130 or @xmath131 is the @xmath132 perfect - fluid four velocit , and @xmath133 is the @xmath132 fluid peculiar velocity potential .
in addition , to avoid a momentum flux relative to @xmath134 , we define the total four velocity @xmath135 as @xcite @xmath136 where the total velocity potential @xmath137 is given by @xcite @xmath138 with @xmath139 , @xmath140 and @xmath141 .
this is the choice of @xmath137 that we will use from now on .
+ thus , the @xmath132 fluid energy - momentum tensor is @xcite @xmath142 where @xmath143 and @xmath144 .
the covariant form of energy - momentum transfer is satisfied for the whole system , while for each component we have @xcite @xmath145 where @xmath146 describe the interaction , @xmath147 for @xmath148 in the late universe , and @xmath149 .
a general @xmath146 relative to the total four velocity can be split as @xcite @xmath150 where @xmath151 is the energy density transfer and @xmath152 is the momentum density transfer rate , relative to @xmath134 . here , we also choose @xmath153 , where @xmath154 is a momentum transfer potential . then , from eq .
( [ decomposef ] ) we find that @xcite @xmath155\,,\\ \label{componentfk } f^{a}_{k}&=&a^{2}\partial_{k}\left(f_{a}+\bar{q}_{a}v\right ) .
\end{aligned}\ ] ] the perturbed energy transfer @xmath156 includes a metric perturbation term @xmath157 and a perturbation @xmath158 .
in addition , we stress that the perturbed momentum transfer @xmath159 is made up of two parts : the momentum transfer potential @xmath160 that arises from energy transport along the total velocity and the intrinsic momentum transfer potential @xmath154 .
hence , the total energy - momentum conservation implies that @xmath161 the general evolution equations for the dimensionless density perturbation @xmath162 is given by @xcite @xmath163v_{a } -3(1+{{w}}_{a}){\dot \phi}&=&\frac{{\delta q}_{a}}{{\bar{\rho}}_{a}}+\nonumber\\ \frac{{\bar{q}}_{a}}{{\bar{\rho}}_{a}}\left[\phi-\delta_{a}-3a{\bar{h}}(c^{2}_{sa}-{w}_{a})v_{a}\right ] , \end{aligned}\ ] ] also the velocity perturbation equation takes the form @xmath164\,,\end{aligned}\ ] ] and the relativistic poisson equation is @xmath165 where @xmath166 is newton s constant .
+ we now consider that @xmath2 does not cluster on sub - hubble scales @xmath167 , and therefore , we could ignore @xmath168 from eq .
( [ contrast1 ] ) . moreover , to avoid the nonphysical sound speed , we choose @xmath169 @xcite .
+ similarly , we also assume that the dynamical effects of the gravitational potential @xmath127 , its time derivative @xmath170 and the transfer of energy between baryons and radiation , may be neglected relative to @xmath1 perturbation , @xmath171 . here
, we also consider the case where @xmath1 component behaves as dust with an eos parameter @xmath172 and with a @xmath1 sound speed @xmath173 .
+ in the linear approach eqs .
( [ contrast1 ] ) and ( [ euler1 ] ) describe the evolution of the @xmath1 perturbation @xmath174 , which can be rewritten as @xcite @xmath175 @xmath176\,.\ ] ] in this linear regime the poisson equation reduces to @xmath177 in order to satisfy the weak equivalence principle and ensure that the particles of the @xmath1 can follow geodesics , we need to impose the following condition @xmath178 here , eq . ( [ euler1 ] ) for @xmath1 component yields @xmath179 this expression means that the @xmath1 velocity perturbation is not affected by the interaction with @xmath2 .
then , we provide a phenomenological covariant choice of the energy exchange four - vector @xmath180 in where one takes @xmath181 we impose the following conditions @xmath182 and @xmath183 to generate @xmath1 cosmic structure formation ( in a forthcoming article we will extend our study , by considering other relations between @xmath184 , @xmath185 and @xmath186 .
it is beyond the scope of the present paper ) , so eq . ( [ parts2_exchange_energy ] ) becomes @xmath187 considering that @xmath1 is more concentrated than the baryon component in the universe , also that @xmath2 does not cluster on sub - hubble scales , using eq .
( [ hubble ] ) and the relation @xmath188 , we have @xmath189 and then , @xmath190 from eqs .
( [ eulerdm ] ) , ( [ poissondm ] ) , ( [ eulerdmfinal ] ) and deriveting eq . ( [ continuitydm ] ) with respect to @xmath44 , we find the evolution of matter density perturbations @xmath191 @xmath192 when @xmath103 , eq .
( [ growth_rate ] ) could be turner into the standard evolution of @xmath1 density perturbations . from this equation
the quantity @xmath6 is an effective gravitational strength ( effective newton constant ) , defined as @xmath193.\ ] ] which can be understood as a self attractive force acting on the @xmath1 density perturbation and quantifies the modifications to gravity due to the effects of @xmath60 and @xmath3 functions .
+ here , we also define the effective hubble friction term as @xmath194 which acts as a frictional force that slows down ( reduces ) the growth of cosmic structure .
+ it is useful to rewrite eq .
( [ growth_rate ] ) in redshift space as @xmath195.\end{aligned}\ ] ] this equation can be solved numerically , considering that @xmath196 where @xmath7 is a some unknown function of @xmath22 so - called the growth index of the linear matter fluctuations .
in the linear regime , it is convenient to define the quantity @xmath197 called the growth factor of @xmath1 density perturbations .
then , eq . ( [ growth_ratez ] ) can be rewritten in function of @xmath198 as @xmath199 this previous equation can be solved numerically , taking into account the condition @xmath200 .
+ in full generality , we define the growth index of @xmath1 perturbations @xmath7 through the following ansatz @xmath201 from eq .
( [ f ] ) , we find that @xmath202\,.\ ] ] now , using eqs .
( [ eofb])-([eofde ] ) , we obtain @xmath203\,.\ ] ] substituting eq .
( [ fb ] ) into eq .
( [ fa ] ) , we get @xmath204\,.\ ] ] let us now to equal eqs .
( [ growth_factor ] ) and ( [ fc ] ) to obtain the general evolution equation for the growth index @xmath7 @xmath205\,.\end{aligned}\ ] ] this equation can be solved numerically by considering the condition that @xmath206 .
+ the parameterization given by eq .
( [ f ] ) is important to simplify rapidly the numerical calculations of eqs .
( [ growth_factor ] ) and ( [ newgrowth_index ] ) .
therefore , the @xmath1 linear growth factor normalized to unity at the present epoch is given by @xmath207 where @xmath22 is the redshift of the universe in which the @xmath1 component dominates the universe ( in this work for convenience used @xmath208 ) .
+ let us stress that by solving numerically eqs .
( [ growth_ratez ] ) and ( [ growth_factor ] ) we can calculate @xmath209 and @xmath198 , respectively .
+ on the other hand , the root - mean - square amplitude of matter density perturbations within a sphere of radius @xmath210 ( being @xmath211 the dimensionless hubble parameter ) is denoted as @xmath212 and its evolution is represented by @xmath213 in where @xmath214 is the normalization of @xmath212 today .
thus , the functions @xmath198 y @xmath215 can be combined to obtain @xmath4 at different redshifts . from here , we obtain @xmath216 the measurements of @xmath4 will be important to constrain different cosmological models .
in this section , we describe how we use the cosmological data currently available to test and constrain the parameter space of our models proposed .
the sne ia data sample used in this work is the join analysis luminous data set ( jla ) @xcite composed by @xmath217 sne with hight - quality light curves . here
, jla data includes several low - redshift samples ( @xmath218 ) , three samples from the sloan digital sky survey sdss - ii at @xmath219 and data from supernova legacy survey ( snls ) in @xmath220 .
+ for the jla data , the observed distance modulus of each sne is modeled by @xmath221 in where @xmath222 and the parameters @xmath223 , @xmath224 and @xmath225 describe the intrinsic variability in the luminosity of the sne , which are derived from the fitting of the light curves . here , @xmath223 is the observed peak magnitude in the rest - frame @xmath226 band , @xmath224 is the stretch measure of the light - curve shape and @xmath225 is the color measure for each sne .
+ on the other hand , the nuisance parameters @xmath227 , @xmath228 , @xmath229 and @xmath230 characterize the global properties of the light - curves of the sne and are estimated simultaneously with the cosmological parameters of interest .
the parameter @xmath227 describes the luminosity of the light - curve , @xmath226 represents the color - luminosity relationships , @xmath229 is the absolute magnitude of the sne in the rest - frame @xmath226 band and @xmath230 denotes the correction of the absolute magnitude @xmath229 with host galaxy properties . from here , we defined @xmath231 @xmath232 where @xmath233 is the host galaxy stellar mass , and @xmath234 is the solar mass .
+ the total covariance matrix for this test is denoted as @xmath235 , and can be written of the following manner @xmath236 where @xmath237 , @xmath238 and @xmath239 denote the diagonal part of the statistical covariance matrix , the statistical covariance matrix and the systematic covariance matrix , respectively .
the details of building of the matrix @xmath235 can be found in @xcite .
@xmath240\,,\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{aligned}\ ] ] where the quantities @xmath241 , @xmath242 and @xmath243 are the covariances of @xmath223 , @xmath224 and @xmath225 for the @xmath244-th sne , respectively , while @xmath245 , @xmath246 , and @xmath247 are the covariances between @xmath223 , @xmath224 and @xmath225 for each @xmath244-th sne .
the terms @xmath248 , @xmath249 and @xmath250 account for the uncertainty in cosmological redshift due to the following quantities : the peculiar velocities , the variation of magnitudes caused by gravitational lensing and the intrinsic variation in sne magnitude , @xcite .
we follow @xcite in using @xmath251 and the prescription suggested by jnsson for @xmath252 @xcite .
the values of @xmath253 are compatible with a constant value of @xmath254 @xcite .
furthermore , @xmath255 denotes the covariance due to a peculiar velocity residual @xcite .
+ on the other hand , the theoretical distance modulus is defined as @xmath256 + 25\;,\ ] ] where the superscript `` @xmath257 '' denotes the theoretical prediction for a sne at a redshift @xmath22 . likewise
, @xmath258 is the luminosity distance , which in a frw cosmology becomes @xmath259 where @xmath260 is the heliocentric redshift , @xmath261 is the cmb rest - frame redshift , `` @xmath262 '' is the speed of the light and @xmath263 represents the cosmological parameters of the model .
considering that @xmath264 , so we rewrite @xmath265 as @xmath266\nonumber\\ & & + 52.385606 - 5\log_{10}(h_{0})\,.\end{aligned}\ ] ] thus , the @xmath267 distribution function for the jla data is @xmath268 where @xmath269 is a column vector of @xmath217 entries of residuals between the theoretical and distance modulus .
@xmath270 is the @xmath271 covariance matrix for all the observed distance modulus reported in @xcite , which contains information over both systematic and statistical errors .
+ rsd data represent a compilation of measurements of the quantity @xmath32 at different redshifts , which were obtained in a model independent way .
these data are apparent anisotropies ( effects ) of the galaxy distribution in redshift space due to the differences of the estimates between the redshifts observed distances and true distances .
they are caused by the component along the line of sight ( los ) of the peculiar velocity of each of the galaxies ( recessional speed ) .
thus , on very small scales ( a few @xmath272 ) ; especially , in the cores of the clustering of galaxies , the peculiar velocities of galaxies are almost randomly oriented such that the structures of the clustering appear elonged along the los when they are viewed in the redshift space ( the " finger of god effect ) @xcite leading to a damping of the clustering of galaxies . by contrast , on large scale ( from a few tens of @xmath272 to @xmath273 ) the observations show that the gravitational growth the galaxies tend to fall towards high - density regions and flow away from low - density regions such that the galaxy clustering in redshift space is enhanced in the los direction in comparison to the transverse direction @xcite .
+ the rsd test is an important probe for distinguishing cosmological @xmath2 models from standard cosmological models such as @xmath14cdm model ; namely , different cosmological models might undergo similar background evolution behavior , but their growth histories of cosmic structures could be distinct in the coupled @xmath2 models .
+ in this work , we utilize the most recent growth rate data derived from redshift space distortions on the pscz , @xmath12df , vvds , @xmath8df , @xmath12mass , boss and wigglez galaxy surveys , and were collected by mehrabi et al .
( see table in @xcite ) .
this sample is used to constrain the free parameters of our theoretical models .
+ the standard @xmath274 for this data set is defined as @xcite @xmath275^{2}}{{\sigma}^{2}(z_{i})}\;\;,\ ] ] where @xmath276 is the observed @xmath277 uncertainty , @xmath278 and @xmath279 represent the theoretical and observational growth rate , respectively .
.summary of rsd data set @xcite . [ cols="^,^,^,^,^,^,^,^",options="header " , ] + + + + +
we constructed a code to calculate numerically the theoretical evolutions of @xmath280 and @xmath198 , respectively , and therefore , the values of @xmath4 , setting functional forms on @xmath60 and @xmath3 such that they can be easily implemented in each of our models .
then , via a markov chain monte carlo ( mcmc ) analysis , we can perform a global fitting in each of them ( listed in table [ bestfits ] ) , by using a combined statistical analysis of cosmic observations such as jla data , the rsd data , the bao data , the cmb given by the planck @xmath13 data and the h data ; from which , we could reduce the uncertainty and put tighter constraints on the values of the cosmological parameters . table [ priors ]
describes the priors used in this work . for each of the models ,
the one - dimension probability contours , the best fitting parameters and their errors ( at @xmath277 and @xmath281 ) are shown in fig .
[ contours ] .
+ the values of the functions @xmath4 , @xmath7 , @xmath60 , @xmath3 , @xmath31 , @xmath6 and @xmath5 evaluated in @xmath282 ( today ) are denoted as @xmath283 , @xmath284 , @xmath285 , @xmath23 , @xmath286 , @xmath287 and @xmath288 , respectively ( see table [ cosmological_state ] . + in the following figs .
the constraints at @xmath277 and @xmath281 on @xmath31 , @xmath30 , @xmath60 , @xmath3 , @xmath4 , @xmath289 , @xmath290 and @xmath7 have been omitted to obtain a better visualization of the results .
+ let us now see fig .
[ behaviours_iq ] , within the coupled models have considered that @xmath291 denotes an energy transfer from @xmath2 to @xmath1 ; on the contrary , @xmath292 denotes an energy transfer from @xmath1 to @xmath2 . in this regard , within
the coupled models have found a change from @xmath291 to @xmath292 and vice versa .
a change of sign on the best reconstructed @xmath60 is linked to the crossing of the non - coupling line , @xmath293 .
table [ crossing_state ] shows the @xmath294 points that satisfy the condition @xmath293 , which were already predicted by eq .
( [ zcrossiq ] ) . moreover , the left below panel in fig .
[ behaviours_iq ] , confirms the statement given by eq .
( [ diqdznull ] ) .
we also verify that if the @xmath22 points satisfy the relation @xmath295 , then they will be different in comparison with the @xmath296 points . according to table [ cosmological_state ] and the upper panels in fig .
[ behaviours_iq ] , note that a non - negligible value of @xmath297 at @xmath277 error is found in the coupled models , and whose order of magnitude is in agreement with the results obtained in @xcite .
however , due to the two minimums obtained in the dr model ( see table [ bestfits ] ) , two different cases ( @xmath298 and @xmath12 ) to reconstruct @xmath299 are worked here . from table
[ bestfits ] we focus on the case @xmath12 , which is in disagreement with the result obtained in eq .
( [ diqdznull7 ] ) ; in this way , the observational data are the fundamental tool to fix the constraints on the cosmological parameters , testing and choosing the possible theoretical models to be worked . on the other hand , from the results presented in fig .
[ behaviours_w ] , we note that in the left and right above panels the universe evolves from the quintessence regime @xmath300 to the phantom regime @xmath301 , and in particular , crosses the phantom divide line @xmath302 @xcite . in the dr model , this crossing feature is more favored with two phantom crossing points in @xmath303 and @xmath304 , respectively , instead , the xcpl model shows only one phantom crossing point in @xmath305 .
likewise , the cpl model also depicts one phantom crossing point in @xmath306 . from these above panels in fig .
[ behaviours_w ] , we also see that in the xcpl model the evolution of @xmath3 is similar to that in the cpl model ; in contrast , the parameter @xmath3 defined in the dr model , starts to evolve from the value @xmath307 during the matter era and reaches the value @xmath308 in the present time .
likewise , a finite value @xmath309 $ ] is obtained in the future .
we stress that there is a significant difference for the evolution of @xmath3 in the xcpl and dr models , and depend on the epoch at which they are compared . from the right above panel in fig .
[ behaviours_w ] , we find that in the dr model when @xmath310 , the amplitude of @xmath3 grows from @xmath311 to @xmath312 .
by contrast , when @xmath313 , the amplitude of @xmath3 decreases from @xmath312 to @xmath314 , whereas in the xcpl model for @xmath315 , the amplitude of @xmath3 decreases more rapidly than that in the dr model .
indeed , these characteristics are a consequence of the reconstructed eos parameters in the cpl , xcpl and dr models , respectively . in addition , in the dr model for the region @xmath316 , @xmath3 deviates significantly from @xmath317 , with a pronounced peak at around @xmath318 and with an average value of @xmath319 .
this behavior is opposite with the evolution of @xmath3 in @xcite .
+ from the left below panel in fig .
[ behaviours_w ] , we also verify that if the @xmath22 points satisfy the relation @xmath302 , then one finds the following condition @xmath320 .
+ the right below panels in figs .
[ behaviours_iq ] and [ behaviours_w ] , show that @xmath60 could take positive or negative values during its evolution from @xmath321 to @xmath322 .
therefore , the values for @xmath31 moves from @xmath323 to @xmath324 , and the values for @xmath30 moves from @xmath325 to @xmath326 , respectively .
these final values for @xmath31 and @xmath30 are indicated in table [ bestfits ] .
+ from the upper panels in figs .
[ behaviours_iq ] and [ behaviours_w ] , and from left above panel in fig .
[ effects1 ] , we focus on the dr model at @xmath310 . here
, @xmath3 grows and @xmath60 could take positive , negative ad null values , and therefore , they will force to the fact that the concentration of @xmath327(@xmath31 ) to grow ( decrease ) more rapidly than those in the @xmath14cdm , cpl and xcpl models , respectively . for @xmath328 ,
@xmath3 decreases and @xmath60 could take positive , negative and null values .
thus , they will induce to the fact that the values obtained for @xmath329 in @xmath282 are closer to those values measured today , with @xmath2 is dominant .
+ considering the right above panel in fig .
[ behaviours_iq ] , the right below panels in figs .
[ behaviours_iq ] and [ behaviours_w ] , and the left above panel in fig .
[ effects1 ] , we see that for @xmath330 the value of the amplitude of @xmath31 ( @xmath30 ) in the dr model is slightly modified by the values of @xmath60 ( @xmath291 and @xmath292 ) relative to the other model , it means that , @xmath60 changes from @xmath291 to @xmath292 , and vice versa . in this model
the amplitude of @xmath31(@xmath30 ) is suppressed ( amplified ) in comparison with those found in the other models .
this result coincides with that found in @xcite . here
, we also confirm that the coincidence problem is alleviated in these coupled models , but they may not solve it . + from table [ cosmological_state ] and from the below panels in fig .
[ effects2 ] , note that the values of @xmath290 deviate significantly from unity in all @xmath22 .
it is in agreement with the resulted found in @xcite .
accordingly , @xmath290 are growing or decreasing functions , and could cross the value @xmath331 at less one time or twice .
furthermore , at @xmath282 , the values of @xmath290 can be roughly larger than @xmath298 ( xcpl model ) or smaller than @xmath298 ( dr model ) .
these observations show the effects of the reconstructions of @xmath60 and @xmath3 on the evolution of @xmath290 in the linear regime .
+ similarly , considering table [ cosmological_state ] , the right above and below panels in fig .
[ effects1 ] and the upper panels in fig .
[ effects2 ] , find the effect of @xmath31 and @xmath60 on the evolution of @xmath289 function . at around @xmath282 ,
the values of @xmath289 in the coupled models are roughly different among them . moreover ,
for @xmath332 the best fitting of @xmath289 in the xcpl model deviates significantly with respect to that obtained in the dr ( cases @xmath298 and @xmath12 ) model .
this is a consequence of the higher quantity of @xmath31 concentrated and of a lesser magnitude of @xmath289 in the xcpl model .
therefore , in the region @xmath333 , the amplitude of @xmath4 is higher and more pronounced in the xcpl with respect to that in the dr model . at the end of the matter era ( at @xmath334 )
the value of @xmath289 in the xcpl model decreases to be approximately smaller than those in the dr and uncoupled models , respectively . in the regime @xmath335 ,
the values of @xmath289 in the dr model become closer to @xmath12 , and thus higher than those found in the xcpl model , reducing the cosmic structure formation ( see right above panel in fig .
[ effects1 ] ) . likewise , according to eqs .
( [ geff ] ) and ( [ heff ] ) an increase on the magnitudes of @xmath60 and @xmath30 tend to amplify the gravitational strength ( @xmath290 ) , but they reduce the magnitude of the frictional force . in fact , an increase in @xmath6 would enhance the growth of structure even at later times .
+ in addition , the right above and right below panels in fig .
[ effects2 ] , show that @xmath289 ( @xmath290 ) during its evolution could take values from @xmath336 to @xmath337 ( from @xmath338 to @xmath339 ) , and therefore the value of @xmath4 could move from @xmath340 to @xmath341 .
this final value for @xmath4 is shown in the right upper panel of fig .
[ effects1 ] , and indicated in table [ cosmological_state ] .
+ from figs .
[ effects1 ] and [ effects2 ] , we consider the evolution of @xmath60 and @xmath30 , finding explicitly that in the xcpl model the functions @xmath290 follows a different behavior from that predicted in the dr model . for this reason ,
the deviations of @xmath290 from standard gravity are significant .
it also explains why @xmath4 is larger in the xcpl model ( and uncoupled models ) than that in the dr model ; especially , when @xmath342 .
more explicitly , for @xmath343 , the values of @xmath290 are close to @xmath298 in the dr model and larger than @xmath298 in the xcpl model , implying the existence of a non - standard gravity .
therefore , in the coupled models the evolution of @xmath60 , @xmath344 , @xmath345 and @xmath3 are different such that their effects can not be ruled out .
the modifications to gravity enhance the structure formation at late times in the xcpl models , but suppresses it in the dr model , when @xmath346 .
in general , the deviation of @xmath290 from unity starts at early times ( @xmath347 ) in the coupled models .
this indicates that the magnitude of @xmath60 is very large there .
thus , for @xmath348 , the substantial difference in the values of @xmath290 is more pronounced . in other words
, in the xcpl model the matter density is much higher than that in the dr model , and therefore affecting more the magnitude of @xmath4 in the xcpl model than that in the dr model .
this explains why the results are very different in this regime , and the differences from uncoupled models are induced mainly by the effective hubble friction term , @xmath289 ( which acts as a frictional force that slows down the linear structure growth ) . + the left above panel in fig .
[ effects3 ] , depicts the evolution of @xmath7 along @xmath22 for the coupled and uncoupled models .
likewise , the right upper and right below panels in fig .
[ behaviours_iq ] , show that @xmath60 could take positive or negative values during its evolution from @xmath349 to @xmath350 , and the values for @xmath31 moves from @xmath351 to @xmath324 . from here , and using the right upper and left below panels in fig .
[ effects3 ] note that the amplitude for @xmath7 is progressively increased to become approximately @xmath352 .
additionally , from the right below panel of this figure and considering the coupled models , note that the values for growth of cosmic structure are very different in the past , and hence the corresponding values for @xmath7 are very closed to zero .
if the values for @xmath4 are progressively increased , then the values for @xmath7 also increase , and become much more stable , when @xmath353 . in table
[ cosmological_state ] show the values of @xmath284 for each of the models studied
. + let us analyze the right upper and left below panels in fig .
[ effects3 ] . from here , we find that the magnitude of @xmath60 has imprinted new physical effects on the evolution of the parameter , @xmath7 . in the dr model
the amplitude of @xmath7 is progressively reduced in the region @xmath354 , with respect to those found in the uncoupled models .
therefore , this shows that the magnitude of @xmath60 is strongly related with the magnitudes of @xmath345 , @xmath4 and @xmath7 , respectively . + we now compare our results with those obtained by other researchers . in @xcite ,
the authors parameterized @xmath7 in terms of the legendre polynomials , and compared it with those obtained from other cosmological models . here
power spectrum data and weak lensing power spectrum data were used .
our results obtained for @xmath7 are very closed to that obtained in the @xmath355 model , at @xmath277 error .
furthermore , in @xcite , the authors provided a convenient analytic formula for @xmath4 , which was applied to different @xmath2 models .
they used rsd data to place observational constraints .
the results obtained by them on @xmath4 are consistent at @xmath277 error with our results .
likewise , pouri et al . in @xcite used the clustering properties of luminous red galaxies ( lrgs ) and the growth rate data to constrain @xmath7 .
the results found by them on @xmath7 and @xmath4 are compatible with our results , at @xmath277 error .
similarly , yang and xu in @xcite , studied a model composed by the cosmological constant , with a nonzero @xmath1 eos parameter .
the result obtained by on @xmath4 is consistent with our result at @xmath277 error .
also , the authors in @xcite , studied the impact of @xmath2 clustering on @xmath7 .
they used two different eos parameters , and found a fitting evolution curve for @xmath4 , which at @xmath277 error is acceptable with our result .
now we summarize our main results : + @xmath356 an analysis combined of data was performed to break the degeneracy among the different cosmological parameters of our models , obtaining constraints more stringent on them . in particular , for the xcpl and dr models ,
the allowed regions for their parameters are significantly reduced by the inclusion of the cmb and rsd data when are compared with studies of models without these data @xcite .
this implies that higher redshift and dynamical probes may be able to discriminate between these models .
+ @xmath356 in the dr model , a novel reconstruction for @xmath3 is proposed here , and whose best fitted value is closed to @xmath25 .
moreover , it has the property of avoiding divergences in a distant future @xmath65 .
this result is consistent with the value predicted by the @xmath14cdm model at @xmath277 error .
likewise , within this coupled scenario , a finite value for @xmath3 is obtained from the past to the future ; namely , the following asymptotic values are found : @xmath357 for @xmath358 , @xmath359 for @xmath360 and @xmath361 $ ] for @xmath65 ( see right above panel in fig . [ behaviours_w ] ) .
therefore , a possible physical description performed by the dr model on the dynamical evolution of @xmath2 should be used to explore its properties .
+ @xmath356 in the coupled models the values of the amplitudes of @xmath30 ( see left upper panel in fig.[effects1 ] ) are slightly modified by the reconstructions of @xmath60 and @xmath3 when they are compared with those in the uncoupled models . nevertheless , they are definitely positive .
this requirement implies that @xmath3 must be always negative in all the cosmic stages of the universe ( see upper panels in fig .
[ behaviours_w ] ) .
+ @xmath356 if @xmath60 takes the values @xmath291 and @xmath292 ( see right upper panel in fig .
[ behaviours_iq ] ) , then the amplitudes of @xmath31 ( @xmath30 ) ( see left upper panel in fig .
[ effects1 ] ) for the two cases in the dr model are smaller ( larger ) in the past than their corresponding @xmath31 ( @xmath30 ) in the xcpl and uncoupled models .
likewise , we also found in the dr model that the values of the amplitudes of @xmath30 ) are significantly affected by the values of both @xmath299 and @xmath3 .
naturally , a smaller proportion of @xmath1 leads to a lesser cosmic structure formation .
therefore , the magnitude of @xmath4 in the dr model is suppressed in comparison with those found in the xcpl and uncoupled models ( see right upper panel in fig . [ effects1 ] ) .
+ @xmath356 for different redshifts , we note that in the coupled models the evolution of @xmath289 and @xmath290 ( see fig .
[ effects2 ] ) follow different behaviors from those found in the uncoupled models .
therefore , they represent a deviation from the evolution predicted by the uncoupled models .
consequently , the dr model predicts an enhancement ( suppression ) on the amplitude of @xmath30 ( @xmath31 ) with respect to that found in the xcpl model ( see left upper panel in fig.[effects1 ] ) .
these effects are significantly sensible to the reconstructions of @xmath60 and @xmath3 , respectively , and decrease when @xmath22 tends to zero .
+ @xmath356 in the coupled @xmath2 models , the decisive role in modifying the cosmic structure formation relative to the uncoupled models is determined mainly by the evolution of @xmath60 , @xmath3 and @xmath31 , respectively . for @xmath282 ,
the values of @xmath4 are very closed to each other . + @xmath356 currently , an enhancement on the amplitude of @xmath4 is the situation revealed in xcpl model when it is compared with that in the dr model , and therefore these scenarios should be considered to study new physical properties of the universe ( see right upper panel in fig.[effects1 ] ) .
+ @xmath356 the behaviors qualitatively presented here show that the plot for @xmath7 has more possibility in discriminating the different coupled @xmath2 models , and therefore @xmath7 could be used to distinguish them ( see left upper panel in fig .
[ effects3 ] ) . +
@xmath364\;,\\ i_{1}(z)&=&\frac{2}{z_{max}}\biggl[\frac{2z}{z_{max}}-\frac{(2+z_{max})}{z_{max}}\ln\biggl(1+z\biggr)\biggr]\;,\\ i_{2}(z)&=&\frac{2}{z_{max}}\biggl[\frac{4 z}{z_{max}}\left(\frac{z}{z_{max}}-\frac{2}{z_{max}}-2\right)+ \left(1+\frac{6.8284}{z_{max}}\right)\left(1+\frac{1.1716}{z_{max}}\right)\ln\biggl(1+z\biggr)\biggr]\;,\\ \tilde{i_{0}}(\tilde{x})&=&\frac{2}{z_{max}}\biggl[\ln{\biggl(1 + 0.5\;z_{max}(1+\tilde{x})\biggr)}\biggr]\;,\\ \tilde{i_{1}}(\tilde{x})&=&{\frac{2}{z_{max}}}\biggl[\biggl(1+\tilde{x}\biggr)-\frac{(2+z_{max})}{z_{max}}\ln{\biggl(1 + 0.5\;z_{max}(1+\tilde{x})\biggr)}\biggr]\;,\\ \tilde{i_{2}}(\tilde{x})&=&{\frac{2}{z_{max}}}\biggl[\biggl(1+\tilde{x}\biggr)\biggl(\tilde{x}-\frac{4}{z_{max}}-3\biggr)+\biggl(1+\frac{6.8284}{z_{max}}\biggr ) \biggl(1+\frac{1.1716}{z_{max}}\biggr)\ln{\biggl(1 + 0.5\;z_{max}(1+\tilde{x})\biggr)}\biggr]\;.\end{aligned}\ ] ]
@xmath367\;,\\ a_{1}&=&\frac{4}{(2+z_{max})^{2}}\biggl[\omega_{1}-\omega_{0}+5\omega_{2}+16\omega_{2}(z_{max}^{-2}+z_{max}^{-1})+2\omega_{1}z_{max}^{-1}\biggr]\;,\\ a_{2}&=&\frac{4}{(2+z_{max})^{2}}\biggl[\omega_{0}+2\omega_{1}-5\omega_{2}-16\omega_{2}(z_{max}^{-2}+z_{max}^{-1})+4\omega_{1}z_{max}^{-1}\biggr]\;,\\ j_{0}&=&\frac{-4}{3(2+z_{max})^{2}}\biggl[\lambda_{0}-\lambda_{1}-\lambda_{2}+2\lambda_{2}(1 + 2z_{max}^{-1})^{2}-2\lambda_{1}z_{max}^{-1}\biggr]\;,\\ j_{1}&=&\frac{-2}{3(2+z_{max})^{2}}\biggl[-\lambda_{0}+\lambda_{1}+\lambda_{2}+4\lambda_{2}(1 + 2z_{max}^{-1})^{2}+2\lambda_{1}z_{max}^{-1}\biggr]\;,\\ j_{2}&=&\frac{-2\sqrt2}{3(2+z_{max})^{2}}\biggl[\lambda_{0}+2\lambda_{1}-\lambda_{2}-4\lambda_{2}(1 + 2z_{max}^{-1})^{2}+4\lambda_{1}z_{max}^{-1}\biggr]\;.\end{aligned}\ ] ] the author is grateful to prof .
f. astorga for his academic support and fruitful discussions in the early stages of this research , and also thank prof .
o. sarbach for useful comments .
this work was in beginning supported by the ifm - umsnh .
conley a et al .
, _ astrophys .
j. suppl . _
* 192 * ( 2011 ) 1 .
jnsson , j. , et al .
2010 , _ mon . not .
* 405 * ( 2010 ) 535 .
betoule m et al . , _ astron . and astrophys . _
* 568 * ( 2014 ) a22 .
j. c. jackson , _ mon . not .
* 156 * ( 1972 ) 1p .
kaiser n. , _ mon . not .
soc . _ * 227 * ( 1987 ) 1 .
a. mehrabi , s. basilakos , f. pace , _ mon . not .
* 452 * ( 2015 ) 2930 - 2939 .
alcock c. and paczynski b. , _ nature .
_ * 281 * ( 1979 ) 358 .
seo , e. r. siegel , d. j. eisenstein , and m. white , _ astrophys .
j. _ * 686 * ( 2008 ) 13y24 .
r. a. battye , t. charnock and a. moss _ phys .
* d 91 * ( 2015 ) 103508 .
l. samushia , et al .
, _ mon . not .
* 439 * ( 2014 ) 3504 .
hudson m. j. , turnbull s. j. , _ astrophys .
j. _ * 751 * ( 2013 ) l30 .
beutler f. , blake c. , colless m. , jones d. h. , staveley - smith l. , et al .
, _ mon . not .
* 423 * ( 2012 ) 3430 .
feix m. , nusser a. , branchini e. , _ phys .
* 115 * ( 2015 ) 011301 .
percival w. j. , et al .
, _ mon . not .
* 353 * ( 2004 ) 1201 .
song y .- s .
, percival w. j. , _ j. cosmol .
phys . _ * 0910 * ( 2009 ) 004 .
tegmark m. et al . , _ phys .
* d 74 * ( 2006 ) 123507 .
guzzo l. et al . ,
_ * 451 * ( 2008 ) 541 .
samushia l. , percival w. j. , raccanelli a. , _ mon . not .
* 420 * ( 2012 ) 2102 .
blake c. et al .
, _ mon . not .
* 415 * ( 2011 ) 2876 ; _ mon . not .
* 418 * ( 2011 ) 1725 .
tojeiro r. , percival w. , brinkmann j. , brownstein j. , eisenstein d. , et al .
, _ mon . not .
* 424 * ( 2012 ) 2339 .
reid b. a. , samushia l. , white m. , percival w. j. , manera m. , et al . , _ mon . not .
* 426 * ( 2012 ) 2719 .
de la torre s. , guzzo l. , peacock j. , branchini e. , iovino a. , et al .
, _ astron .
astrophys . _
* 557 * ( 2013 ) a54 .
planck 2015 results , xiii .
cosmological parameters , _ astron .
astrophys . _
* 594 * ( 2016 ) a13 .
wmap collaboration , g. hinshaw et al .
, _ astrophys .
_ * 208 * ( 2013 ) 19 .
f. beutler et al . , _ mon . not .
roy . astron .
* 416 * ( 2011 ) 3017 .
a. j. ross et al . , _ mon . not .
soc . _ * 449 * ( 2015 ) 835 .
w. j. percival et al . , _ mon . not .
* 401 * ( 2010 ) 2148 .
e. a. kazin et al . , _ astrophys .
j. _ * 710 * ( 2010 ) 1444 .
n. padmanabhan et al . , _ mon . not .
* 427 * ( 2012 ) 2132 . c. h. chuang and y. wang , _ mon . not .
_ * 435 * ( 2013 ) 255 .
chuang and y. wang , _ mon . not .
* 433 * ( 2013 ) 3559 .
l. anderson et al . , _ mon . not .
_ * 441 * ( 2014 ) 24 .
e. a. kazin et al . , _ mon . not .
* 441 * ( 2014 ) 3524 .
t. delubac et al . , _ astron .
_ * 574 * ( 2015 ) a59 .
a. font - ribera et al . , _ j. cosmol .
phys . _ * 05 * ( 2014 ) 027 .
d. j. eisenstein , w. hu , _ astrophys .
j. _ * 496 * ( 1998 ) 605 .
d. j. eisenstein et al . , _ astrophys .
j. _ * 633 * ( 2005 ) 560 .
m. d. p. hemantha , y. wang and c - h .
, _ mon . not .
* 445 * ( 2014 ) 3737
. j. r. bond , g. efstathiou and m. tegmark , _ mon . not .
_ * 291 * ( 1997 ) l33 .
w. hu and n. sugiyama , _ astrophys .
j. _ * 471 * ( 1996 ) 542 .
j. neveu , v. ruhlmann - kleider , p. astier , m. besanon , j. guy , a. mller , e. babichev , _
arxiv:1605.02627v1_. g. s. sharov , _ j. cosmol
. astropart .
phys . _ * 06 * ( 2016 ) 023 . c. zhang et al .
astrophys . _
* 14 * ( 2014 ) 1221 .
j. simon , l. verde and r. jimenez , _ phys .
* d 71 * ( 2005 ) 123001 .
m. moresco et al . , _ j. cosmol .
* 8 * ( 2012 ) 006 .
e. gastaaga , a. cabre , l. hui , _ mon . not .
soc . _ * 399 * ( 2009 ) 1663 .
a. oka et al . , _ mon . not .
* 439 * ( 2014 ) 2515 . c. blake et al . , _ mon . not .
_ * 425 * ( 2012 ) 405 .
d. stern , r. jimenez , l. verde , m. kamionkowski and s. a. stanford , _
j. cosmol .
phys . _ * 02 * ( 2010 ) 008 .
m. moresco , _ mon . not .
* 450 * ( 2015 ) l16-l20 .
n. g. busca et al . , _ astron .
astrophys . _
* 552 * ( 2013 ) a96
. p. j. e. peebles and b. ratra , _ astrophys .
j. _ * 325 * ( 1988 ) l17 .
p. j. e. peebles and b. ratra , _ rev .
* 75 * ( 2003 ) 559 .
v. sahni , _ lect .
notes phys . _
* 653 * ( 2004 ) 141 .
e. j. copeland , m. sami and s. tsujikawa , _ int .
* d 15 * ( 2006 ) 1753 .
s. weinberg , _ rev .
phys . _ * 61 * ( 1989 ) 1 .
v. sahni and a. a. starobinsky , _ int .
_ * d 9 * ( 2000 ) 373 .
u. seljak , et al .
* d 71 * ( 2005 ) 103515 .
e. rozo et al . , _ astrophys .
j. _ * 708 * ( 2010 ) 645 .
r. r. caldwell , _ phys .
* b 545 * ( 2002 ) 23 ; s. nojiri and s. d. odintsov , _ phys .
lett . _ * b 562 * ( 2003 ) 147 ; r. gannouji , d. polarski , a. ranquest and a. a. starobinsky , _ j. cosmol
. astropart .
phys . _ * 09 * ( 2006 ) 016 ; x. cheng , y. gong and e. n. saridakis , _ j. cosmol
. astropart .
phys . _ * 04 * ( 2009 ) 001 .
e. elizalde , s. nojiri , and s. d. odintsov _ phys .
* d 70 * ( 2004 ) 043539 ; z. k. guo , y. s. piao , x. m. zang and y. z. zhang , _ phys .
* b 608 * ( 2005 ) 177 .
b. ratra , and p. j. e. peebles , _ phys .
* d 37 * ( 1988 ) 3406 ; k. coble , s. dodelson , and j. a. frieman , _ phys .
_ * d 55 * ( 1997 ) 1851 ; r. r. caldwell , r. dave , and p. j. steinhardt , _ phys .
* 80 * ( 1998 ) 1582 . c. armendariz - picon , v. mukhanov , p. j. steinhardt , _ phys .
_ * 85 * ( 2000 ) 4438 , _ phys .
* d 63 * ( 2001 ) 103510 ; t. chiba , t. okabe , m. yamaguchi , _ phys .
rev . _ * d 62 * ( 2000 ) 023511 .
a. y. kamenshchik , u. moschella and v. pasquier , _ phys .
* b 511 * ( 2001 ) 265 ; m. c. bento , o. bertolami and a. a. sen , _ phys .
* d 66 * ( 2002 ) 043507 ; m. k. mak and t. harko , _ phys .
* d 71 * ( 2005 ) 104022 ; m. r. garousi , m. sami , s. tsujikawa , _ phys .
* b 606 * ( 2005 ) 1 ; m. r. garousi , m. sami , and s. tsujikawa , _ phys .
* d 71 * ( 2005 ) 083005 .
m. s. turner , _ phys .
* d 28 * ( 1983 ) 1243 .
k. a. malik , d. wands , and c. ungarelli , _ phys .
_ * d 67 * ( 2003 ) 063516 .
r. cen , _ astrophys .
j. _ * 546 * ( 2001 ) l77 ; m. oguri , k. takahashi , h. ohno and k. kotake , _ astrophys . j. _ * 597 * ( 2003 ) 645 . z. k. guo , n. ohta , and s. tsujikawa , _ phys .
* d 76 * ( 2007 ) 023508 ; l. amendola , g. c. campos , r. rosenfeld , _ phys
_ * d 75 * ( 2007 ) 083506 .
he and b. wang , _ j. cosmol .
phys . _ * 06 * ( 2008 ) 010 . c. g. bhmer , g. caldera - cabral , r. lazkoz , and r. maartens , _ phys .
rev . _ * d 78 * ( 2008 ) 023505 .
j. valiviita , e. majerotto and r. maartens , _ j. cosmol
. astropart .
* 07 * ( 2008 ) 020
. s. campo , r. herrera and d. pavon , _ j. cosmol
. astropart .
_ * 01 * ( 2009 ) 020 .
g. caldera - cabral , r. maartens and b. m. schaefer , _ j. cosmol
. astropart .
phys . _ * 07 * ( 2009 ) 027 . c. g. bhmer , g. caldera - cabral , n. chan , r. lazkoz and r. maartens , _ phys .
* d 81 * ( 2010 ) 083003 .
l. p. chimento , _ phys .
_ , * d 81 * ( 2010 ) 043525 . e. abdalla , l. r. abramo and j. c. c. de souza , _ phys . rev . _ * d 82 * ( 2010 ) 023508 .
r. g. cai and q. su , _ phys .
* d 81 * ( 2010 ) 103514 .
j. h. he , b. wang , and e. abdalla , _ phys . rev . _ * d 83 * ( 2011 ) 063515 .
s. cao , n. liang and z. h. zhu , _ int .
* d 22 * ( 2013 ) 1350082 .
y. h. li and x. zhang , _ eur .
j. c. _ * 71 * ( 2011 ) 1700 .
d. pavn , w. zimdahl , _ phys .
* b 628 * ( 2005 ) 206 ; d. pavn , b. wang , _ gen.rel.grav .
_ * 41 * ( 2009 ) 1 - 5 .
b. wang , y. g. gong and e. abdalla , _ phys .
* b 624 * ( 2005 ) 141 .
b. wang , c. y. lin and e. abdalla , _ phys .
* b 637 * ( 2006 ) 357 .
s. del campo , r. herrera , g. olivares , and d. pavn , _ phys .
* d 74 * ( 2006 ) 023501 .
f. cueva solano and u. nucamendi , _ j. cosmol
. astropart .
_ * 04 * ( 2012 ) 011 ; f. cueva solano and u. nucamendi , _ arxiv : 1207.0250 _ * 07 * ( 2012 ) 02 .
w. zimdahl , _ int . j. mod
_ * d 14 * ( 2005 ) 2319 .
s. das , p. s. corasaniti , and j. khoury , _ phys .
* d 73 * ( 2006 ) 083509 .
g. huey and b. d. wandelt , _ phys .
* d 74 * ( 2006 ) 023519 .
b. wang , j. zang , c. y. lin , e. abdalla , and s. micheletti , _ nucl .
* b 778 * ( 2007 ) 69 .
a. r. cooray and d. huterer , _ astrophys . j. _ * 513 * ( 1999 ) l95 .
m. chevallier , d. polarski , _ int . j. mod .
_ * d 10 * ( 2001 ) 213 ; e. v. linder , _ phys .
* 90 * ( 2003 ) 091301 .
j. barboza , e. m. and j. alcaniz , _ phys .
* b 666 * ( 2008 ) 415 .
e. m. barboza jr .
et al ; _ phys .
rev . _ * d 80 * ( 2009 ) 043521 .
q. j. zhang and y. l. wu , _ j. cosmol .
_ * 08 * ( 2010 ) 038 .
h. li and x. zhang , _ phys .
* b 703 * ( 2011 ) 119 ; j. z. ma and x. zhang , _ phys .
* b 699 * ( 2011 ) 233 .
r. a. daly and s. djorgovski , _ astrophys .
j. _ * 597 * ( 2003 ) 009 .
d. huterer and a. cooray , _ phys .
* d 71 * ( 2005 ) 023506 .
a. shafieloo , u. alam , v. sahni and a. a. starobinsky , _ mon . not .
_ * 366 * ( 2006 ) 1081 .
a. hojjati , l. pogosian and g. b. zhao , _ j. cosmol .
phys . _ * 04 * ( 2010 ) 007 . o. sarbach and m. tiglio , _ liv .
rel . _ * 15 * ( 2012 ) 9 .
e. f. martinez and l. verde , _ j. cosmol .
phys . _ * 08 * ( 2008 ) 023 . sargent w. l. w. and turner e. l. , _ astrophys . j. _ * 212 * ( 1977 ) 3 .
hamilton a. j. s. , _ `` the evolving universe '' .
kluwer academic _
( 1998 ) 185 - 275 .
peacok j. a. et al .
, _ nature . _
* 410 * ( 2001 ) 169 .
hawkins e. et al . , _ mon . not .
soc . _ * 346 * ( 2003 ) 78 .
a. de felice , r. kase and s. tsujikawa , _ phys .
* d 83 * ( 2011 ) 043515 .
s. a. appleby and e. v. linder , _
j. cosmol .
phys . _ * 08 * ( 2012 ) 026 . a. de felice and s. tsujikawa , _ j. cosmol
. astropart .
phys . _ * 03 * ( 2012 ) 025 .
nishimichi t. and oka .
a. , _ mon . not .
* 444 * ( 2014 ) 1400 - 1418 .
matsubara t. and suto y. , _ astrophys .
j. _ * 470 * , l1 ( 1996 ) .
ballinger w. e. peacok j. a. and heavens a. f. , _ mon . not .
* 282 * ( 1996 ) 877 .
kazuya koyama , roy maartens , and yong - seon song , _ j. cosmol .
* 10 * ( 2009 ) 017 ; p.
brax , c. van de bruck , d. f. mota , n. j. nunes , and h. a. winther , _ phys .
* d 82 * ( 2010 ) 083503 .
s. nesseris and l. perivolaropoulos , _ j. cosmol
. astropart .
phys . _ * 01 * ( 2007 ) 018 .
e. majerotto , j. valiviita and r. maartens , _ mon . not .
* 402 * ( 2010 ) 2344 .
t. clemson , k. koyama , g. b. zhao , r. maartens and j. valiviita _ phys .
* d 85 * ( 2012 ) 043007 .
s. nesseris , a. de felice , and s. tsujikawa , _ phys .
* d 82 * , ( 2010 ) 124054 ; j. lu , _ phys .
* b 680 * , ( 2009 ) 404 ; j. neveu , et al .
, _ astron . and astrophys . _
* 555 * ( 2013 ) a53 .
l. amendola , _ phys .
* d 69 * , ( 2004 ) 103524 .
a. bueno b.,j .
gracia - bellido , and d. sapone , _ j. cosmol .
* 10 * ( 2011 ) 01 .
s. tsujikawa , a. de felice and j. alcaniz , _ j. cosmol
. astropart .
phys . _ * 01 * ( 2013 ) 030 .
a. pouri , s. basilakos , m. plionis , _ j. cosmol
. astropart .
phys . _ * 08 * ( 2014 ) 042 .
w. yang and l. xu , _ phys .
* d 89 * ( 2014 ) 083517 . | based on general relativity ( gr ) we consider two different cosmological scenarios in where reconstruct the energy exchange ( @xmath0 ) between cold dark matter ( @xmath1 ) fluid and dark energy ( @xmath2 ) fluid , which is modelled with a @xmath2 varying equation of state ( eos ) parameter @xmath3 .
we here investigate the main cosmological effects on the growth rate of matter density perturbations ( @xmath4 ) , on the effective hubble friction term ( @xmath5 ) , on the effective newton constant ( @xmath6 ) and on the growth index of the linear matter fluctuations ( @xmath7 ) .
our study demonstrates that in the coupled models the evolution of these quantities are modified with respect to the predictions in the uncoupled models , and therefore could be used to distinguish among coupled @xmath2 scenarios .
finally , we also perform a combined statistical analysis using current observational data ( geometric and dynamical probes ) to put more stringent constraints on the parameters space of the cosmic scenarios studied . |
low energy supersymmetry ( susy)@xcite has several features that make it very promising as a class of models for beyond the standard model ( sm ) of particle physics@xcite . the minimal supergravity ( msugra)@xcite is a well - studied model that greatly reduces the number of parameters of the minimal supersymmetric standard model ( mssm)@xcite through a few simple requirements .
the universal input parameters given at the unification scale are : ( i ) the gaugino mass parameter @xmath10 , ( ii ) the scalar mass parameter @xmath11 , and ( iii ) the trilinear susy breaking parameter @xmath2 .
the model that uses radiative breaking of electroweak symmetry requires another parameter @xmath1 which is the ratio of higgs vacuum expectation values .
additionally the sign of the higgsino mixing parameter @xmath12 is an input .
with r - parity@xcite assumed to be preserved , the model has an attractive dark matter@xcite candidate , namely the lightest neutralino @xmath0 which is the lightest supersymmetric particle ( lsp ) for most of the parameter space . in a scenario where @xmath0 is the lsp ,
the parameter space with other sparticles lighter than @xmath0 are ignored .
the model is typically associated with a bino - dominated lsp that leads to over - abundance of dark matter .
reduction of the relic density to satisfy the wmap data can be achieved if i ) there is coannihilation of lsp with another sfermion ( usually stau @xmath13 or rarely @xmath14 ) that has mass close to the mass of lsp@xcite , ii ) there is appropriate mixture of bino and higgsinos in the composition of the lsp so that there may be coannihilating charginos in the lsp-@xmath15 annihilation@xcite , the so called focus point@xcite / hyperbolic branch@xcite region , iii ) the lsp is sufficiently small in mass and sfermions are light so that light sfermion exchange may enhance the lsp - lsp annihilation rates , or iv ) there is a possibility of having s - channel higgs exchanges where the exchanges may occur via cp - odd higgs boson @xmath16 or via cp - even heavy ( light ) higgs bosons @xmath17 ( @xmath18 ) leading to the `` funnel region '' of dark matter satisfied zone in the @xmath19 plane@xcite .
most recently xenon100@xcite experiment has published an updated result for the exclusion limit of @xmath20 for direct search of dark matter for a weakly interacting massive particle ( wimp ) mass of 50 gev at 90% confidence limit . additionally , the recent 5-tower result of the cdms - ii@xcite experiment published a result of 2 signal events with @xmath21 events as background . as mentioned by the authors
, there is a significant chance that these two events might have been caused by background rather than by any real signal .
however , a low recoil of 12.3 to 15.5 kev@xcite possibly suggests a lighter wimp with a mass not quite far from 100 gev .
we may thus assume that the spin - independent direct detection cross section limit ( @xmath22 ) set by the experiment , namely @xmath23 @xmath24 at a wimp mass in the vicinity of 70 gev may become an interesting zone in the near future in a light lsp scenario .
several analyses have used this theme in the recent past@xcite and there have been several recent works@xcite in susy models in the context of the cdms - ii result . for practical purposes , we will probe the region of @xmath25 to @xmath26 @xmath24 in a lighter lsp scenario ( not too far from 100 gev ) that satisfies the following relic density bounds from wmap@xcite : @xmath27 a possibly preferred scenario where the lsp is light and at the same time is dominated by bino may be achieved if neutralino relic density is brought to an acceptable level ( from a typical over - abundance ) via a light higgs pole annihilation mechanism .
computation of relic density in msugra with nonvanishing trilinear coupling for large @xmath1 showed such a possibility in ref.@xcite that also imposed constraints from flavor physics like that from @xmath28 and @xmath29 . here in this work
, we will particularly explore the reach of the direct detection cross section in relation to any small recoil data of cdms - ii in such a scenario with light higgs pole annihilation .
in addition to imposing the lep2 limit@xcite for @xmath30 , the mass of the light higgs boson , we will also explore the parameter space in regard to the constraints like @xmath28 and @xmath29 processes .
we will particularly discuss the effect of considering nonminimal flavor violating ( non - mfv ) scenarios with a wider point of view that may be considered without affecting much the spectra or the dark matter related results .
we will further probe the possible signatures coming out of lhc in regard to mssm higgs bosons and weakly interacting gauginos @xmath31 in the dark matter satisfied zone with larger direct detection cross section @xmath22 .
we note that lhc may help to probe the mass of the lighter higgs boson ( @xmath30 ) from @xmath32 mode , whereas an existence of light higgs pole annihilation would indicate a light lsp scenario with a value of @xmath33 that would satisfy @xmath34 .
additionally , at the lhc , one would expect to see a signal of three leptons and a missing transverse energy namely @xmath35 mainly from the pair production of @xmath36 followed by their decays via leptonic modes , @xmath37 as a classic signature of susy .
we also note that in regard to the decay @xmath38 , the endpoint of _ opposite sign same flavor _ ( ossf )
dilepton invariant mass distribution @xmath39 leads to a good determination of the mass difference @xmath40@xcite .
this article is organized as follows . in sec . 2
, we discuss the issues of direct detection of dark matter particularly the spin - independent cross section @xmath22 in relation to the parameter space of msugra that satisfies the wmap data via s - channel higgs pole annihilation . in sec . 3 , we identify the preferred sign of the trilinear coupling in relation to the above s - channel annihilation . in sec . 4 , we discuss the implications of wmap data and low energy constraints on the msugra parameter space that is interesting for a cdms - ii type of event with a small lsp mass . in sec . 5
, we discuss several possible signatures of this scenario at the lhc .
we also comment on how the correlations between different lhc signatures and the direct detection of dark matter may help determine the masses of @xmath41 and @xmath42 .
finally , we summarise our findings in sec .
in general , the results of cross section of lsp annihilation to quarks can be used to find the elastic scattering cross section of the lsp with quark , thanks to crossing symmetry .
direct detections of lsps involve measurement of recoil energy of a nucleus due to lsp - nucleon scattering@xcite .
typical detector materials are scintillators like nai , semiconductors like ge and noble liquids like xe .
the elastic scattering of neutralino with nucleons are divided into two types ( i ) spin - independent ( si ) : a neutralino coherently interacts with the nucleus and ( ii ) spin - dependent ( sd ) : a neutralino interacts with matter via axial vector coupling .
the effective lagrangian that describes @xmath43 elastic scattering for a small velocity is given by , @xmath44 here , the first term represents spin - dependent scattering while the second term refers to spin - independent scattering .
the terms assume summing over the quark flavors @xmath45 as well as that for up and down type of quarks ( for @xmath46 and @xmath47 respectively ) .
the neutralino - quark coupling coefficients @xmath48 and @xmath49 contain all susy model dependent information@xcite .
the spin - independent scattering cross section of a neutralino with a target nucleus of proton number ( atomic number ) @xmath50 and neutron number @xmath51 ( @xmath16 being the mass number ) is given by , @xmath52^{2}~. \label{sitotal}\ ] ] here , @xmath53 is the reduced mass defined by @xmath54 , where @xmath55 refers to the mass of the nucleus .
the quantities @xmath56 and @xmath57 contain all the information of short - distance physics and nuclear partonic strengths .
these are given by , @xmath58 @xmath59 and @xmath60 refer to interactions of neutralino with quarks and gluons ( via quark / squark loop diagrams ) respectively .
for @xmath59 one has , @xmath61 the quantities @xmath62 may be computed from a few hadronic mass data .
the gluon related part namely @xmath63 is given by@xcite , @xmath64 the numerical values of @xmath59 may be seen in ref.@xcite .
we compute @xmath65 by using the code darksusy@xcite .
in contrast to @xmath65 , the spin - dependent cross section denoted by @xmath66 ( for scattering of lsp with the target nucleus ) does not depend on @xmath16 or @xmath50 , rather it scales with @xmath67 where @xmath68 is the total nuclear spin . in general , the spin - independent neutralino - nucleon scattering cross sections ( where @xmath69 ) are appreciably smaller than the corresponding spin - dependent cross sections ( @xmath70 ) .
however considering the fact that @xmath71 and @xmath72 one finds that @xmath65 to be considerably larger for moderately heavy elements ( @xmath73 ) like xe , ge etc .
the experiments like edelweiss@xcite and cdms@xcite use natural germanium ( almost purely spinless ) as the target material and these hardly have any spin - sensitivities@xcite .
it is important to know the composition of the @xmath74 in terms of bino , wino and higgsinos in mssm while analyzing the wmap satisfied relic density region as well as the lsp - nucleon scattering cross section for a direct detection .
the amount of higgsino mixing in a typically bino - dominated lsp scenario of msugra depends on the relative value of @xmath75 and @xmath76 ( the mass of bino ) .
the scalar cross section depends on t - channel higgs exchange ( @xmath77 ) and s - channel squark exchange diagrams . unless , the squark masses are close to the mass of the lsp
, the higgs exchange diagrams usually dominate over the s - channel diagrams@xcite .
we focus on a decoupling region of higgs@xcite , so that there is a considerable difference of masses between @xmath18 and @xmath17-bosons .
we also focus on the case where lsp is principally composed of bino , @xmath1 is large and there is a resonance annihilation via @xmath18-boson that is consistent with the relic density from the wmap limits .
following the results of refs.@xcite we see that both the contributions involving lighter as well as heavier higgs bosons in the direct detection are important . indeed in the characteristic zone of the parameter space
as mentioned above , the heavier higgs boson contribution becomes dominant in comparison to the contribution from the lighter higgs boson@xcite .
this holds in spite of the fact that @xmath78 can be larger by a factor of 3 to 4 compared to @xmath30 . in this light higgs annihilation zone
we will see that @xmath22 can be large enough to reach the cdms - ii limit .
in this analysis we especially investigate the parameter space that satisfies the relic density limits from wmap data via lsp annihilations through light higgs resonance . this necessitates a closer study of the region with small @xmath10 near its domain just allowed by the lep2 limit of lighter chargino mass @xmath79@xcite .
we will see how the lep2 constraint affects the lowest possible value of @xmath10 in a nonvanishing @xmath2 scenario .
this in turn requires a discussion on the behaviors of i ) @xmath12 and ii ) @xmath80 for a variation of @xmath2 .
we note that in msugra the variation of @xmath12 with @xmath2 has the following pattern . for @xmath81 ,
@xmath75 increases with @xmath82 , whereas @xmath75 increases with @xmath2 as long as @xmath83 . on the other hand , @xmath12 is near its minimum when one has @xmath84 . as a result , one finds that for a given set of @xmath11 , @xmath10 and @xmath1 a change of @xmath2 from @xmath85 to @xmath86 where @xmath87 causes @xmath75 to vary from a large to a larger value .
this in turn causes @xmath15 to be dominated by wino in the region with a smaller @xmath10 , the amount of wino content is larger for the case of @xmath86 in comparison to that of @xmath85 where the same content is still significantly large .
thus , while probing the region of small gaugino masses with wino - dominated @xmath15 , it is important to examine the variation of @xmath80 with nonvanishing @xmath2 . in this context ,
table [ tabgauginomasses ] shows the masses of electroweak gauginos , the value of higgsino mass parameter , the masses of lighter chargino , lighter neutralino and h - boson for @xmath88 , @xmath89 gev , @xmath90 gev and @xmath91 for @xmath92 gev .
this will indeed demonstrate the effect of two - loop rges along with threshold corrections on the electroweak gaugino masses .
clearly , both @xmath76 and @xmath80 are smaller for @xmath85 in comparison to the same quantities corresponding to @xmath86 . on the other hand ,
the values of @xmath12 for both signs of @xmath2 are sufficiently large so that the lighter chargino as well as the lighter neutralino are both highly gaugino - dominated .
the value of @xmath30 is smaller for @xmath93 .
thus , as we decrease @xmath10 , the lep2 bound of chargino mass is reached at a relatively larger @xmath10 for @xmath85 in comparison to the case of @xmath86 .
similarly , the lep2 limit of @xmath30 is reached at a relatively larger @xmath10 for @xmath85 than the case of @xmath86 .
the latter favors larger loop corrections from the top - stop sector@xcite . in brief ,
imposition of lep2 bounds of @xmath79 and @xmath30 ( considering also the aforesaid 3 gev theoretical uncertainty ) would allow us to reach to a smaller @xmath10 value for a negative @xmath2 than for a positive @xmath2@xcite . as a result
, bino can be light enough to satisfy a light higgs - pole annihilation condition @xmath34 while obeying the wmap limits in a lep2 satisfied parameter zone .
such a light lsp would undoubtedly be interesting in view of the cdms - ii results with two possible events .
.comparison of spectra electroweak gaugino masses , higgsino masses , the masses of lighter chargino , lighter neutralino and h - boson for @xmath88 , @xmath89 gev , @xmath90 gev and @xmath91 for @xmath92 gev .
this is given as an example to demonstrate the effect of two - loop rges along with threshold corrections on the electroweak gaugino masses which are smaller for @xmath93 .
these are of course not valid parameter points satisfying wmap data for neutralino relic density . [ cols="^,^,^,^,^,^,^,^,^,^,^,^ " , ] in this trilepton signal i.e. , @xmath94 channel , any choice of the three charged leptons ( @xmath95 @xmath96 or @xmath12 ) must consist of an ossf pair and an additional lepton .
the end point of the ossf dilepton invariant mass @xmath97 distribution leads to a good determination of the mass difference @xmath98 @xcite . on the other hand , after reconstruction of the light higgs boson mass @xmath30 from @xmath18-decays and assuming @xmath18-pole annihilation as the mechanism for satisfying the wmap data , it is possible to estimate the mass @xmath99 , which is @xmath100
. then the mass of @xmath42 may be estimated by combining @xmath99 and with the edge in the @xmath101 distribution .
it may be recalled that for the present choice of benchmark points , @xmath41 is completely bino - dominated , while @xmath42 is wino - dominated .
this leads to one to one correspondence between @xmath102 soft gaugino mass parameter @xmath76 with @xmath99 and @xmath103 soft gaugino mass parameter @xmath80 with @xmath104 .
hence , we expect that the knowledge of both @xmath99 and @xmath104 may be used to determine the mass parameters @xmath105 and @xmath80
. it would be interesting to study whether the mass pattern of @xmath105 and @xmath80 could hint towards the gaugino mass unification which model like msugra assumes .
finally , we conclude this section , by highlighting following important observations : 1 .
the light higgs boson mass can be determined with a reasonably good precision via @xmath106 decay mode .
2 . in the present scenario , where @xmath107 is large ,
both neutral higgs boson masses can be determined via @xmath108 decay .
both @xmath109 and @xmath110 or equivalently @xmath76 and @xmath80 can be estimated by combining the data from the direct detection of dark matter and the lhc .
4 . as mentioned before , our choice of parameter space prefers light gluinos and also lighter third generation sleptons and squarks .
this naturally leads to large production rate for gluinos as well as lighter stop and sbottom squarks which in turn give rise to large number of multi @xmath111-jet final states in association with large missing energy both at 7 and 14 tev lhc run .
this will be another striking signature of this scenario .
recent experiments like xenon and cdms - ii provide us with exclusion limits within 3 - 4 @xmath112 for wimp mass in the vicinity of 50 to 70 gev .
cdms - ii results show a 2-signal events with @xmath21 events as background .
it may be possible that these events did not come from any real signal . in spite of this , in case we do not see the events as just background , a low recoil of 12.3 to 15.5 kev could possibly indicate a light wimp detection whose mass could not be too far away than 100 gev .
a large direct detection cross section where lsp is light ( and that also satisfies the wmap data ) in a generic model like msugra can be possible in a parameter space with a large @xmath1 and nonvanishing trilinear coupling @xmath2 .
the relic density constraint is satisfied via h - pole resonant annihilation mechanism . regarding direct detection @xmath3 cross section which reaches the level of the cdms - ii events , both the contributions involving the lighter as well as the heavier higgs bosons are important . in this large @xmath1 ( @xmath113 )
based scenario with bino - dominated @xmath0 , where light higgs pole annihilation brings down the relic density to an acceptable level , the heavier higgs boson contribution indeed dominates .
this is in spite of the fact that @xmath78 can be larger by a factor of 3 to 4 compared to @xmath30 . unlike the past analyses referred in this paper regarding the light higgs - pole dominated lsp pair - annihilation processes for the cases with large values of @xmath1 , here we have worked on the direct detection prospects of the same region in relation to the recent interest of cdms - ii results .
we have further shown in adequate detail the preference of only one sign of @xmath2 that gives feasibility to obtain the desired zone of light - higgs pole annihilation for dark matter .
we particularly perform an open analysis by considering non - mfv assumptions in relation to the @xmath114 and @xmath29 constraints , unlike most other recent works .
we have considered the possibility of general flavor violation on @xmath29 , specifically keeping in mind about the recent analyses as referred in this work . we believe that unless the mfv - based limits from flavor physics are violated by a very large extent , it is important not to discard appropriate scenarios that keep the region with the light higgs - pole dominated pair - annihilation active
. this is particularly relevant because of the ongoing and upcoming precision direct detection dark matter experiments and the advantage of having an associated light higgs spectra as well as a relatively light overall spectra . in the last half of the paper we present a qualitative discussions on the interplay between collider searches for mssm higgs bosons @xmath115 , as well as some other sparticles at the lhc and direct detection of dark matter for the aforementioned parameter space . in particular , we discuss the possibility of determining both @xmath116 and @xmath117 by combining the data from the direct detection of dark matter experiment with that of the edge in the @xmath118 distribution and the light higgs boson mass measurement at the lhc .
we further note that because of @xmath1 being large , copious production of @xmath119 in association with the @xmath111-quark would be possible at the lhc .
such an @xmath120 mass can be reconstructed by looking at the @xmath121 decay mode .
we also point out that large @xmath1 and relatively low @xmath10 values of our benchmark points lead to a large number of multi - b jets in association with missing energy signal .
dd thanks p2i , cnrs for the support received as a post - doctoral fellow .
dkg and mm acknowledge support from the department of science and technology , india under grants sr / s2/hep-12/2006 and sr / mf / ps-03/2009-vb - i respectively .
dkg also thanks theory division , cern for the hospitality when part of this work was done . for reviews on supersymmetry ,
see , _ eg _ , h. p. nilles , ( 110 , 1 , 1984 ) ; h. e. haber and g. kane , ( 117 , 75 , 1985 ) ; j. wess and j. bagger , _ supersymmetry and supergravity _ , 2nd ed . , ( princeton , 1991 ) ; m. drees , p. roy and r. m. godbole , _ theory and phenomenology of sparticles _ , ( world scientific , singapore , 2005 ) . a. h. chamseddine , r. arnowitt and p. nath , phys . rev . lett . * 49 * , 970 ( 1982 ) ; r. barbieri , s. ferrara and c. a. savoy , phys . lett .
b * 119 * , 343 ( 1982 ) ; l. j. hall , j. lykken and s. weinberg , phys .
d * 27 * , 2359 ( 1983 ) ; p.
nath , r. arnowitt and a. h. chamseddine , nucl .
b * 227 * , 121 ( 1983 ) ; n. ohta , prog .
. phys . * 70 * , 542 ( 1983 ) ; for reviews see @xcite and p. nath , r. arnowitt and a.h .
chamseddine , _ applied n = 1 supergravity _ ( world scientific , singapore , 1984 ) .
g. bertone , d. hooper and j. silk , phys .
rept . * 405 * , 279 ( 2005 ) ; g. bertone and d. merritt , mod .
a * 20 * , 1021 ( 2005 ) ; j. carr , g. lamanna and j. lavalle , rept .
phys . * 69 * , 2475 ( 2006 ) .
g. jungman , m. kamionkowski and k.greist , phys . rep . * 267 * , 195 ( 1996 ) .
j. r. ellis , t. falk and k. a. olive , phys .
b * 444 * , 367 ( 1998 ) ; j. r. ellis , t. falk , k. a. olive and m. srednicki , astropart .
* 13 * , 181 ( 2000 ) [ erratum - ibid . *
15 * , 413 ( 2001 ) ] ; a. lahanas , d. v. nanopoulos and v. spanos , phys . rev .
d * 62 * , 023515 ( 2000 ) ; r. arnowitt , b. dutta and y. santoso , nucl .
b * 606 * , 59(2001 ) ; t. nihei , l. roszkowski and r. ruiz de austri , jhep * 0207 * , 024 ( 2002 ) ; v. a. bednyakov , h. v. klapdor - kleingrothaus and v. gronewold , phys .
d * 66 * , 115005 ( 2002 ) .
j. edsjo and p. gondolo , phys .
d * 56 * , 1879 ( 1997 ) .
s. mizuta and m. yamaguchi , phys .
b * 298 * , 120 ( 1993 ) . j. l. feng , k. t. matchev and t. moroi , phys .
d * 61 * , 075005 ( 2000 ) ; phys .
* 84 * , 2322 ( 2000 ) ; j. l. feng , k. t. matchev and f. wilczek , phys .
b * 482 * , 388 ( 2000 ) ; j. l. feng and f. wilczek , phys . lett .
b * 631 * , 170 ( 2005 ) ; u. chattopadhyay , t. ibrahim and d. p. roy , phys .
d * 64 * , 013004 ( 2001 ) ; u. chattopadhyay , a. datta , a. datta , a. datta and d. p. roy , phys .
b * 493 * , 127 ( 2000 ) ; s. p. das , a. datta , m. guchait , m. maity and s. mukherjee , eur . phys .
j. c * 54 * , 645 ( 2008 ) , [ arxiv:0708.2048 [ hep - ph ] ] .
m. drees and m. nojiri , phys . rev .
* d47 * , 376 ( 1993 ) ; r. arnowitt and p. nath , phys .
lett . * 70 * , 3696 ( 1993 ) ; h. baer and m. brhlik , phys . rev . *
d53 * , 597 ( 1996 ) , phys .
d * 57 * , 567 ( 1998 ) ; h. baer , m. brhlik , m. diaz , j. ferrandis , p. mercadante , p. quintana and x. tata , phys . rev . * d63 * , 015007 ( 2000 ) ; j. r. ellis , t. falk , g. ganis , k. a. olive and m. srednicki , phys .
b * 510 * , 236 ( 2001 ) ; a. b. lahanas and v. c. spanos , eur .
j. * c23 * , 185 ( 2002 ) .
e. aprile _ et al .
_ [ xenon100 collaboration ] , phys . rev . lett .
* 105 * , 131302 ( 2010 ) .
z. ahmed _ et al .
_ [ the cdms - ii collaboration ] , science * 327 * , 1619 ( 2010 ) .
i. gogoladze , r. khalid , s. raza and q. shafi , arxiv:0912.5411 [ hep - ph ] ; m. asano , s. matsumoto , m. senami and h. sugiyama , arxiv:0912.5361 [ hep - ph ] ; m. farina , d. pappadopulo and a. strumia , phys .
b * 688 * , 329 ( 2010 ) ; x. g. he , t. li , x. q. li , j. tandean and h. c. tsai , phys .
b * 688 * , 332 ( 2010 ) ; m. ibe and t. t. yanagida , phys .
d * 81 * , 035017 ( 2010 ) ; m. kadastik , k. kannike , a. racioppi and m. raidal , arxiv:0912.3797 [ hep - ph ] .
a. bottino , f. donato , n. fornengo and s. scopel , phys .
d * 81 * , 107302 ( 2010 ) ; m. holmes and b. d. nelson , phys .
d * 81 * , 055002 ( 2010 ) ; k. cheung and t. c. yuan , phys .
b * 685 * , 182 ( 2010 ) ; j. hisano , k. nakayama and m. yamanaka , phys .
b * 684 * , 246 ( 2010 ) ; d. p. roy , phys .
d * 81 * , 057701 ( 2010 ) ; r. allahverdi , b. dutta and y. santoso , phys .
b * 687 * , 225 ( 2010 ) ; d. feldman , z. liu and p. nath , phys .
d * 81 * , 095009 ( 2010 ) ; j. cao , k. i. hikasa , w. wang , j. m. yang and l. x. yu , jhep * 1007 * , 044 ( 2010 ) ; d. feldman , z. liu , p. nath and g. peim , phys .
d * 81 * , 095017 ( 2010 ) ; d. feldman , z. liu and p. nath , phys .
d * 81 * , 117701 ( 2010 ) .
e. komatsu _ et al .
_ [ wmap collaboration ] , astrophys . j. suppl . *
180 * , 330 ( 2009 ) , arxiv:0803.0547 [ astro - ph ] .
a. djouadi , m. drees and j. l. kneur , phys .
b * 624 * , 60 ( 2005 ) .
i. hinchliffe , f. e. paige , m. d. shapiro , j. soderqvist and w. yao , phys .
d * 55 * , 5520 ( 1997 ) [ arxiv : hep - ph/9610544 ] ; v. d. barger and c. kao , phys .
d * 60 * , 115015 ( 1999 ) [ arxiv : hep - ph/9811489 ] .
j. r. ellis , a. ferstl and k. a. olive , phys .
b * 481 * , 304 ( 2000 ) .
m. w. goodman and e. witten , phys .
d * 31 * , 3059 ( 1985 ) ; k. greist , phys .
d * 38 * , 2357 ( 1988 ) ; j. ellis and r. flores , nucl . phys .
b * 307 * , 883 ( 1988 ) ; r. barbieri , m. frigeni and g. giudice , nucl .
phys b * 313 * , 725 ( 1989 ) ; a. bottino et .
al . , phys .
b * 295 * , 330 ( 1992 ) ; m. drees and m. m. nojiri , phys .
d * 48 * , 3483 ( 1993 ) ; m. drees and m. nojiri , phys . rev .
d * 47 * , 4226 ( 1993 ) and phys .
d * 48 * , 3483 ( 1993 ) ; m. a. shifman , a. i. vainshtein and v. i. zakharov , phys . lett .
b * 78 * , 443 ( 1978 ) ; a. i. vainshtein , v. i. zakharov and m. a. shifman , usp . fiz .
nauk * 131 * , 537 ( 1980 ) .
p. gondolo , j. edsjo , p. ullio , l. bergstrom , m. schelke and e. a. baltz , jcap * 0407 * , 008 ( 2004 )
. v. sanglard _ et al . _
[ the edelweiss collaboration ] phys .
d * 71 * , 122002 ( 2005 ) [ arxiv : astro - ph/0503265 ] .
v. a. bednyakov and h. v. klapdor - kleingrothaus , phys . part .
nucl . * 40 * , 583 ( 2009 ) [ arxiv:0806.3917 [ hep - ph ] ] .
m. drees and m. nojiri , phys .
d * 48 * , 3483 ( 1993 ) .
haber and y. nir , phys .
* b306 * 327 ( 1993 ) ; h.e .
haber , hep - ph/9505240 ; a. dobado , m.j .
herrero and s. penaranda , eur .
j. * c17 * 487 ( 2000 ) ; j.f .
gunion and h.e .
haber , phys .
* d67 * 075019 ( 2003 ) .
g. belanger , f. boudjema , a. cottrant , a. pukhov and a. semenov , nucl .
b * 706 * , 411 ( 2005 ) .
for the latest limits on the sparticle masses from lep experiments : see http://lepsusy.web.cern.ch/lepsusy/ j. m. frere , d. r. t. jones and s. raby , nucl .
b * 222 * , 11 ( 1983 ) ; j. a. casas , a. lleyda and c. munoz , nucl . phys .
b * 471 * , 3 ( 1996 ) .
m. s. carena and h. e. haber , prog .
* 50 * , 63 ( 2003 ) .
u. chattopadhyay , d. das , a. datta and s. poddar , phys .
d * 76 * , 055008 ( 2007 ) .
g. degrassi , s. heinemeyer , w. hollik , p. slavich and g. weiglein , eur .
j. c * 28 * , 133 ( 2003 ) ; b. c. allanach , a. djouadi , j. l. kneur , w. porod and p. slavich , jhep * 0409 * , 044 ( 2004 ) ; s. heinemeyer , hep - ph/0408340 ; int . j. mod .
a * 21 * , 2659 ( 2006 ) ; s. heinemeyer , w. hollik and g. weiglein , phs . rep . * 425*,265,2006 . a. l. kagan and m. neubert , eur .
j. c * 7 * , 5 ( 1999 ) ; k. g. chetyrkin , m. misiak and m. munz , phys .
b * 400 * , 206 ( 1997 ) [ erratum - ibid .
b * 425 * , 414 ( 1998 ) ] ; p. gambino and m. misiak , nucl .
b * 611 * , 338 ( 2001 ) .
m. ciuchini , g. degrassi , p. gambino and g. f. giudice , nucl .
b * 527 * , 21 ( 1998 ) ; m. ciuchini , g. degrassi , p. gambino and g. f. giudice , nucl .
b * 534 * , 3 ( 1998 ) ; g. degrassi , p. gambino and g. f. giudice , jhep * 0012 * , 009 ( 2000 ) .
s. chen _ et al .
_ [ cleo collaboration ] , phys . rev .
lett . *
87 * , 251807 ( 2001 ) ; b. aubert _ et al .
_ [ babar collaboration ] , [ arxiv : hep - ex/0207076 ] ; p. koppenburg _ et al . _ [ belle collaboration ] , phys .
lett . * 93 * , 061803 ( 2004 ) .
s. r. choudhury and n. gaur , phys .
b * 451 * , 86 ( 1999 ) ; k. s. babu and c. kolda , phys .
lett . *
84 * , 228 ( 2000 ) ; a. dedes , h. k. dreiner , u. nierste , and p. richardson , phys .
lett . * 87 * , 251804 ( 2001 ) ; r. arnowitt , b. dutta , t. kamon and m. tanaka , phys .
b * 538 * ( 2002 ) 121 ; j. k. mizukoshi , x. tata and y. wang , phys .
d * 66 * , 115003 ( 2002 ) ; s. baek , p. ko , and w. y. song , jhep * 0303 * , 054 ( 2003 ) ; t. ibrahim and p. nath , phys .
d * 67 * , 016005 ( 2003 ) ; j.r .
ellis , k.a .
olive and v.c .
spanos , phys .
b * 624 * , 47 ( 2005 ) ; t. aaltonen _ et al .
_ [ cdf collaboration ] , phys . rev .
lett . * 100 * , 101802 ( 2008 ) .
a. djouadi , m. drees and j. l. kneur , jhep * 0603 * , 033 ( 2006 ) . for a recent article on mfv
see for example : g. colangelo , e. nikolidakis and c. smith , eur .
j. c * 59 * , 75 ( 2009 ) . | we consider the prospect of lightest neutralino ( @xmath0 ) as a dark matter candidate in the light of recent interesting observations from the xenon100 and cdms - ii experiments in minimal supergravity framework with large @xmath1 and nonvanishing @xmath2 . within the wmap satisfied zone
, there is a large direct detection reach of lighter @xmath0 in the lighter higgs boson mediated resonance annihilation domain of the above scenario .
it is seen that the heavier higgs boson plays a dominating role in the @xmath3 cross section in the associated zone of parameter space in spite of having a larger mass .
possible lhc signatures are discussed .
-1 cm 16.5 cm 21.5 cm -0.5 cm # 1 # 1 ( # 1,#2,#3)*#1 * , # 2 ( # 3)#1#2#3#4#1 * # 2 * , # 3 ( # 4 ) # 1#2#3jhep * # 1 * , # 3 ( # 2 ) # 1#2#3jhep * # 1 * , # 3 ( # 2 ) # 1#2#3phys .
rev .
d * # 1 * , # 3 ( # 2 ) # 1#2#3phys .
rev .
d * # 1 * , # 3 ( # 2 ) # 1#2#3phys .
rev .
lett . *
# 1 * , # 3 ( # 2 ) # 1#2#3phys .
rev .
lett . * # 1 * , # 3 ( # 2 ) # 1#2#3nucl .
phys .
b * # 1 * , # 3 ( # 2 ) # 1#2#3nucl . phys .
b * # 1 * , # 3 ( # 2 ) # 1arxiv : hep - ph/#1 ( # 1,#2,#3)phys .
rep .
( # 1,#2,#3 ) ( # 1,#2,#3)eur .
phys .
j. c ( # 1,#2,#3 ) 0_b_0 22 2 ^ 2_w 2m_z^2 2b2 v#1v_#1 # 1 ^ 0_#1 # 1m_#1 # 1m_#1 # 1m_^0_#1 # 1m_^+_#1 2wsec^2_w 2(g-2 ) _ = 11 # 1#2 # 1#1| # 1| # 1 # 1#1 # 1#1 versim#1#2 2_2^+ 2_2 ^ - 2 _ 2^ 2 m _ 2 ^ - # 1#2 * probing the light higgs pole resonance annihilation of dark matter in the light of xenon100 and cdms - ii observations .
+ * 0.5 cm utpal chattopadhyay@xmath4 , debottam das@xmath5 , dilip kumar ghosh@xmath4 and manas maity@xmath6 + @xmath7 _ department of theoretical physics , indian association for the cultivation of science , + 2a & b raja s.c .
mullick road , jadavpur , kolkata 700 032 , india _
+ @xmath8 _ laboratoire de physique thorique , umr 8627 , universit de paris - sud 11 & cnrs , + btiment 210 , 91405 orsay cedex , france _ + @xmath9 _ department of physics , visva - bharati , santiniketan 731235 , india _ |
let @xmath6 denote the set of positive integers . in this paper graphs
are assumed to be finite , undirected and without loops , though they may contain multiple edges .
graphs having no multiple edges will be called simple . if @xmath1 is a graph , then for a vertex @xmath7 @xmath8 denotes the degree of @xmath9 in @xmath1 .
moreover , let @xmath10 and @xmath11 denote the maximum and minimum degrees of vertices in @xmath1 , respectively .
a vertex is defined to be isolated in @xmath1 , if its degree is zero .
if @xmath12 is a subgraph of @xmath1 , then we say that @xmath12 covers ( misses ) a vertex @xmath9 of @xmath1 , if @xmath13 ( @xmath14 ) .
a subgraph is spanning , if it covers all the vertices of the graph . a point that should be made clear here ,
is that if a vertex @xmath9 of @xmath1 is not a vertex of a subgraph @xmath12 , then we assume that @xmath14 .
the length of a path @xmath15 of a graph @xmath1 is the number of edges lying on @xmath15 .
if @xmath16 are non - negative integers , then a subgraph @xmath0 of a graph @xmath1 with @xmath17 is called an @xmath18$]-factor of @xmath1 if for any vertex @xmath19 of @xmath1 @xmath20 .
a subset @xmath21 of edges of a graph @xmath1 is called matching , if @xmath22 is a @xmath23$]-factor of @xmath1 .
clearly , matchings can be defined as a set of edges that contain no adjacent edges .
usually , a vertex that is ( not ) incident to an edge from a matching , is said to be covered ( missed ) by the matching .
a matching is maximum , if it contains as many edges as possible , while it is perfect , if any vertex is incident to an edge from the matching .
a @xmath2-edge - coloring of a graph @xmath1 is a partition of @xmath24 into @xmath2 matchings .
usually , these matchings into which @xmath24 is partitioned , are called color - classes of the edge - coloring . the least integer @xmath2
for which @xmath1 has a @xmath2-edge - coloring is called the chromatic index of @xmath1 and is denoted by @xmath25 .
clearly , @xmath26 for any graph @xmath1 , and the following classical theorems of shannon and vizing give non - trivial upper bounds for @xmath25 : ( shannon @xcite ) .
for every graph @xmath1@xmath27.\ ] ] ( vizing , @xcite ) : @xmath28 , where @xmath29 denotes the maximum multiplicity of an edge in @xmath1 .
note that shannon s theorem implies that if we consider a cubic graph @xmath1 , then @xmath30 , thus @xmath31 can take only two values . in 1981
holyer proved that the problem of deciding whether @xmath32 or not for cubic graphs @xmath1 is np - complete holyer , thus the calculation of @xmath33 is already hard for cubic graphs . the quantitative aspect of the investigation of maximum @xmath2-edge - colorable subgraphs of graphs and particularly , @xmath34-regular graphs has attracted a lot of attention , previously .
the basic problem that researchers were interested was the following : what is the proportion of edges of a graph ( or an @xmath34-regular graph , and particularly , cubic graph ) , that we can cover by its @xmath2 matchings ?
for the case @xmath35 in @xcite an investigation is carried out in the class of simple cubic graphs , and in @xcite for the general case .
let us also note that the relation between @xmath36 and @xmath37 has also been investigated in the regular graphs of high girth @xcite .
the same is true for the case @xmath38 .
albertson and haas investigate these ratios in the class of simple cubic and @xmath39-regular graphs in @xcite , and steffen investigates the problem in the class of bridgeless cubic graphs in @xcite .
similar investigations are done in @xcite for subcubic graphs . in @xcite
the problem is addressed in the class of cubic graphs .
finally , a best - possible bound is proved in @xcite for the case @xmath40 in the class of all graphs . however , it worths to be mentioned that the quantitative line of the research was not the only one .
previously , a special attention was also paid to structural properties of maximum @xmath2-edge - colorable subgraphs , and sometimes this kind of results have helped researchers to get quantitative results .
a typical example of a structural result is the one proved in @xcite , which states that in any cubic graph @xmath1 there is a maximum @xmath41-edge - colorable subgraph @xmath0 , such that the graph @xmath42 is @xmath41-edge - colorable .
recently , in @xcite new such results are presented for maximum @xmath4-edge - colorable subgraphs of graphs @xmath1 .
particularly , there it is shown that any set of vertex - disjoint cycles of a graph @xmath1 ( particularly , any @xmath41-factor ) can be extended to a maximum @xmath4-edge - colorable subgraph of @xmath1 if @xmath43 .
also , there it is shown that for any maximum @xmath4-edge - colorable subgraph @xmath0 of @xmath1 @xmath44 for each @xmath45 , where @xmath46 is the set of edges of a graph @xmath47 with exactly one end - vertex in @xmath48 .
finally , in @xcite it is shown that the edges of a cubic graph lying outside a maximum @xmath49-edge - colorable subgraph form a matching .
though this result does not have a direct generalization , using the ideas of the proof of vizing theorem for simple graphs from @xcite , in @xcite it is shown that a simple graph @xmath1 has a maximum @xmath4-edge - colorable subgraph @xmath0 , such that the edges of @xmath1 that do not belong to @xmath0 form a matching . in this paper
, we concentrate on spanning maximum @xmath2-edge - colorable subgraphs of graphs . in the beginning of the paper
we introduce three different parameters that are related with the mentioned problems .
we first show that these parameters are the same .
this enables us to introduce the fourth parameter @xmath3 that equals to the common value of the mentioned three parameters .
we show that @xmath4 is an upper bound for @xmath3 , and then we proceed with the characterization of graphs with @xmath5 .
finally , we relate @xmath3 to some well - known graph - theoretic parameters .
non - defined terms and concepts can be found in @xcite .
we start with the introduction of the three main parameters of the paper .
if @xmath1 is a graph without isolated vertices , then define : @xmath50\textrm{-factor}\}.\ ] ] @xmath51 @xmath52since a maximum @xmath2-edge - colorable subgraph is a @xmath2-edge - colorable subgraph , and a @xmath2-edge - colorable subgraph is a @xmath53$]-factor , we have that for any graph without isolated vertices , one has : @xmath54 let @xmath55 be a spanning @xmath2-edge - colorable subgraph .
consider all maximum @xmath2-edge - colorable subgraphs of @xmath1 , and among them choose the ones that cover maximum possible number of vertices . from these subgraphs , choose a subgraph @xmath56 such that @xmath57 is maximized .
let us show that @xmath56 is a spanning subgraph . on the opposite assumption , consider a vertex @xmath58 missed by @xmath56 . consider the vertices @xmath59 ( @xmath60 ) that are adjacent to @xmath58 . since @xmath56 is a maximum @xmath2-edge - colorable subgraph of @xmath1 , we have : now if @xmath69 , then define a subgraph @xmath70 of @xmath1 as follows : @xmath71clearly @xmath70 is a maximum @xmath2-edge - colorable subgraph of @xmath1 .
moreover , @xmath70 covers more vertices of @xmath1 than @xmath56 does , which contradicts the choice of @xmath56 .
thus ( c ) must hold .
we are ready to complete the proof of the theorem .
since @xmath55 is a spanning @xmath2-edge - colorable subgraph , there is an edge @xmath72 . by ( b ) , we have @xmath73 , thus there is an edge @xmath74 such that @xmath75 . consider a subgraph @xmath76 of @xmath1 as follows : @xmath77clearly @xmath76 is a maximum @xmath2-edge - colorable subgraph of @xmath1 .
due to ( c ) , @xmath76 covers maximum possible number of vertices , like @xmath56 .
however , @xmath78 which contradicts the choice of @xmath56 . the proof of the lemma [ continuousspectra ] is completed .
@xmath79 by the above - mentioned lemma , we already have that @xmath81 .
thus , it suffices to show that if @xmath1 is a graph without isolated vertices , then @xmath82 .
suppose that @xmath83 , and let @xmath0 be a @xmath53$]-factor of @xmath1 .
let @xmath84 be a sub - forest of @xmath0 with @xmath85 .
clearly , @xmath84 is a spanning subgraph of @xmath1 . since @xmath84 is @xmath86-edge - colorable and @xmath87 , we have that @xmath88 . @xmath79 ( tutte @xcite ) a graph @xmath1 has a perfect matching , if and only if for any @xmath93 one has @xmath94 , where for a graph @xmath0 @xmath95 denotes the number of components of @xmath0 that contain odd number of vertices .
now , let us characterize the class of graphs with @xmath5 . clearly , if @xmath98 are components of @xmath1 , then @xmath99 .
thus , a graph @xmath1 satisfies the equality @xmath5 if and only if some of its components satisfies the same equality .
this observation enables us to focus on the characterization of connected graphs @xmath1 that satisfy @xmath5 .
since @xmath1 is not a tree , it must contain a cycle .
let us assume that @xmath1 contains at least two cycles , and let @xmath101 be an edge of @xmath1 lying on a cycle of @xmath1 . observe that : @xmath102taking into account that @xmath5 , we have that @xmath103 . since @xmath104 is connected and @xmath105 , we have that @xmath104 is either a tree or an odd cycle .
now , if @xmath104 is a tree , then @xmath1 must be unicyclic @xcite , which we assumed to be not the case .
hence @xmath104 is an odd cycle .
however , this case is also impossible since if @xmath104 is an odd cycle , then @xmath106 and @xmath107 , and therefore @xmath108 , which contradicts the choice of @xmath1 .
we conclude that @xmath1 is unicyclic .
let us show that any degree one vertex of @xmath1 is adjacent to a vertex of @xmath109 .
on the opposite assumption , we can consider a vertex @xmath58 of @xmath1 such that @xmath111 and @xmath58 is adjacent to @xmath112 vertices of degree one .
let @xmath113 be the degree one neighbours of @xmath58 , and let @xmath19 be the other neighbour of @xmath58 .
observe that since @xmath1 is not a tree , @xmath19 is not of degree one .
let @xmath114 be the component of @xmath115 containing the vertex @xmath19 .
clearly , @xmath109 is a cycle of @xmath114 . we need to consider two cases .
+ case 1 : @xmath116 . in this case , we have that @xmath117 and @xmath118 , hence @xmath108 , which contradicts the choice of @xmath1 .
+ case 2 : @xmath119 .
since @xmath114 is connected , @xmath114 contains a cycle and @xmath120 , we have that @xmath121 .
hence @xmath122 , since @xmath123 , which contradicts the choice of @xmath1 .
+ the considered two cases imply that any degree one vertex of @xmath1 is adjacent to a vertex of @xmath109 .
observe that this implies that all vertices of @xmath1 that are at least of degree two lie on @xmath109 .
we are ready to complete the proof of the lemma .
for this purpose we consider the following two cases , and in each of them we exhibit a contradiction .
+ case 1 : @xmath1 contains two degree two vertices that are adjacent .
let @xmath58 and @xmath19 be adjacent degree two vertices of @xmath1 , and let @xmath124 and @xmath125 be the other ( @xmath126 and @xmath127 ) neighbours of @xmath58 and @xmath19 , respectively . consider the graph @xmath128 obtained from @xmath1 by removing the vertices @xmath58 and @xmath19 , and adding an edge connecting @xmath124 and @xmath125 .
since @xmath128 is connected and @xmath129 , we have that @xmath130 .
let @xmath131 be a spanning @xmath132-edge - colorable subgraph of @xmath128 .
consider a subgraph @xmath0 of @xmath1 obtained from @xmath131 as follows : @xmath133it is easy to see that @xmath0 is a spanning @xmath132-edge - colorable subgraph of @xmath1 , hence @xmath134 contradicting the choice of @xmath1 .
+ case 2 : @xmath1 contains no two degree two vertices that are adjacent .
observe that this case includes the case when there are no degree two vertices in @xmath1 .
for each degree two vertex @xmath58 of @xmath1 choose the edge @xmath135 incident to @xmath58 such that @xmath136 is the next neighbour of @xmath58 in the direction of clockwise circumvention of @xmath109 , and let @xmath137 be the matching of @xmath1 that contains all such edges @xmath135 .
consider a subgraph @xmath0 of @xmath1 obtained as follows : all edges of @xmath1 that are incident to a degree one vertex add to @xmath0 , and add @xmath137 to @xmath0 , too .
clearly , @xmath0 is a spanning @xmath132-edge - colorable subgraph of @xmath1 , hence @xmath134 contradicting the choice of @xmath1 . + the proof of the lemma [ spdeltatreecycle ] is completed .
@xmath79 the proved lemma implies that in order to characterize the connected graphs @xmath1 with @xmath138 , we can focus on trees . for this purpose for an arbitrary tree @xmath84
we introduce the following two sets : @xmath139 we will give a method for the construction of such a subgraph .
we start with @xmath145 .
consider the following partition of vertices of @xmath84 : @xmath146now , add all edges @xmath147 to @xmath0 , such that @xmath148 and @xmath149 . observe that for any @xmath150 one has @xmath151 since @xmath152 has one neighbour in @xmath153 .
after this , remove all edges that we have added to @xmath0 and the vertices incident to them from @xmath84 .
repeat this process until @xmath154 becomes empty or @xmath155 .
it can be easily seen that the components of the resulting subgraph @xmath0 of @xmath84 are stars , such that their centers are of degree at most @xmath156 .
hence @xmath0 is @xmath142-edge - colorable .
moreover , it meets the requirements of the lemma .
[ spdeltacor2]if @xmath84 is a tree with @xmath157 and the subgraph @xmath0 from lemma [ treedeltaminus1 ] does not cover @xmath19 , then there is a spanning @xmath86-edge - colorable subgraph @xmath131 of @xmath84 , such that @xmath159 .
now , we introduce an operation let will help us to characterize the trees @xmath84 with @xmath158 .
let @xmath160 be a tree with @xmath161 , and let @xmath162 be a star with @xmath163 .
consider the tree @xmath164 obtained from @xmath160 and @xmath162 by identifying a degree one vertex of @xmath162 with a vertex @xmath165 .
first , we establish some properties of the operation @xmath166 .
let @xmath172 . clearly ,
suppose that the tree @xmath84 has been obtained from @xmath160 and @xmath162 , by identifying the vertices @xmath174 , and the degree one vertex @xmath175 .
moreover , let @xmath176 be the center of @xmath162 .
\(a ) since @xmath177 , then @xmath178 . let us show that @xmath179 .
as @xmath174 , corollary [ spdeltacor ] implies that there is a @xmath180-edge - colorable subgraph @xmath181 of @xmath160 , such that @xmath182 . consider the subgraph @xmath0 of @xmath84 obtained from @xmath181 by adding @xmath183 to it .
clearly , @xmath0 is @xmath142-edge - colorable subgraph of @xmath84 , hence @xmath184 .
\(b ) clearly , @xmath178 .
let us show that @xmath185 .
take a spanning @xmath186-edge - colorable subgraph @xmath181 of @xmath160 .
consider the subgraph @xmath0 of @xmath84 obtained from @xmath181 by adding @xmath187 to it .
clearly , @xmath0 is a spanning @xmath186-edge - colorable subgraph of @xmath84 .
hence @xmath188 .
\(c ) let us show that @xmath189 .
take a spanning @xmath186-edge - colorable subgraph @xmath181 of @xmath160 .
consider the subgraph @xmath0 of @xmath84 obtained from @xmath181 by adding @xmath187 to it .
clearly , @xmath0 is a spanning @xmath190-edge - colorable subgraph of @xmath84 .
hence @xmath191 .
observe that @xmath195 , as otherwise @xmath196 and hence all edges of @xmath162 would have to be colored , which would mean that @xmath197 .
this implies that @xmath181 is a spanning @xmath2-edge - colorable subgraph of @xmath160 , hence @xmath198 , which contradicts our assumption .
@xmath79 if @xmath84 is a star then clearly @xmath158 . on the other hand , if @xmath84 is obtained from @xmath160 with @xmath199 by applying lemma [ operprop](d ) , then by lemma [ operprop](d ) , @xmath84 satisfies @xmath158 .
now , assume that @xmath84 satisfies @xmath158 .
if @xmath84 is a star then we are done .
otherwise , assume that @xmath84 is not a star .
then there is a vertex @xmath176 of @xmath84 , that is of degree @xmath163 , such that @xmath176 is adjacent to exactly @xmath190 vertices of degree one .
let @xmath160 be the tree obtained from @xmath84 by removing the vertex @xmath176 and all its neighbours that are of degree one .
moreover , let @xmath152 be the vertex of @xmath160 such that @xmath200 .
let us show that @xmath164 .
clearly , it suffices to show that @xmath174 .
suppose that @xmath201 , that is @xmath202 .
then , clearly , @xmath203 .
consider a spanning subgraph @xmath0 of @xmath84 obtained from any spanning @xmath204-edge - colorable subgraph of @xmath160 by adding all edges incident to @xmath176 except @xmath205 .
it is not hard to see that @xmath0 is @xmath206-edge - colorable , hence @xmath207 contradicting the choice of @xmath84 .
let us note that the statement of the last corollary for bridgeless cubic graphs first appeared in the proof of theorem 4.1 from @xcite .
however , an attentive reader probably has already realized that the proof given in @xcite is wrong .
retaining the notations of @xcite , let us , first explain , what is wrong there .
the gap is that when the author removes the edges @xmath211 and @xmath212 from a maximum @xmath41-edge - colorable subgraph @xmath0 and adds the edges @xmath213 and @xmath214 to it to get a new maximum @xmath41-edge - colorable subgraph @xmath215 , he may leave the other ( @xmath216 and @xmath217 , respectively ) end - vertices isolated , so after this operation one can not conclude that @xmath218 as it is done there .
let us note that this bound is tight , since any regular graph without a perfect matching achieves it .
it can be shown that this bound can be improved by one if @xmath1 is non - regular ( that is , @xmath231 ) .
however , we will not prove this , because below we will prove a significantly better bound for @xmath3 .
[ yuliutheorem ] @xcite : let @xmath232 . then a graph @xmath1 has an @xmath18$]-factor , if and only if for all @xmath93 @xmath233 , where @xmath234 is the number of vertices of degree @xmath235 in the graph @xmath236 . by theorem [ yuliutheorem ] it suffices to show that for each @xmath93 @xmath238 , where @xmath239 is the number of isolated vertices of @xmath236 .
observe that by tutte - berge formula , we have : @xmath240 @xmath79 let us note that there are examples of graphs such that the difference between the upper bound offered by theorem [ deltadeltaratio ] and @xmath3 is arbitrarily big . to see this ,
let @xmath0 be an @xmath34-regular graph containing a perfect matching @xmath248 .
consider a graph @xmath1 obtained from @xmath0 by replacing one edge of @xmath248 by a path of length three .
observe that @xmath1 contains a perfect matching , hence @xmath92 , however the bound offered by theorem [ deltadeltaratio ] is @xmath249 . in theorem [ nubound
] , we have shown that an upper bound for @xmath3 is provable in terms of the difference between @xmath250 and @xmath97 .
it is natural to wonder , whether such a bound is possible to prove in terms of the ratio of @xmath250 and @xmath97 .
the following proposition shows the impossibility of such a bound .
let @xmath252 be any positive integer with @xmath253 .
set : @xmath254 and @xmath255 .
consider the tree @xmath1 obtained from a path of length @xmath256 and the star @xmath257 by joining the center of the star to one of end - vertices of the path .
observe that : @xmath258 , @xmath259 and @xmath260 .
clearly , we have that @xmath251 . @xmath79 d. aslanyan , v. v. mkrtchyan , s. s. petrosyan , g. n. vardanyan , on disjoint matchings in cubic graphs : maximum @xmath41- and @xmath49-edge - colorable subgraphs , under review , ( available at : http://arxiv.org/abs/0909.2767 ) | a subgraph @xmath0 of a graph @xmath1 is called spanning , if any vertex of @xmath1 is not isolated in @xmath0 , while it is called maximum @xmath2-edge - colorable , if @xmath0 is @xmath2-edge - colorable and contains as many edges as possible .
we introduce three parameters that are related with spanning @xmath2-edge - colorable subgraphs .
our first result states that these parameters are the same .
this allows us to introduce the fourth parameter @xmath3 that equals to the common value of the mentioned three parameters .
we show that @xmath4 is an upper bound for @xmath3 , and then we offer a characterization of graphs with @xmath5 .
finally , we prove some bounds for @xmath3 that involve well - known graph - theoretic parameters . |
recent cosmological observations support that the universe is experiencing an accelerated expansion , and that the transition to the accelerated phase has been realized in the recent cosmological past @xcite . in order to explain this remarkable behavior , and despite the intuition that this can be achieved only through a fundamental theory of nature , physicists can still propose some paradigms for its description , such are theories of modified gravity @xcite , or `` field '' models of dark energy .
the field models that have been discussed widely in the literature consider a canonical scalar field ( quintessence ) @xcite , a phantom field , that is a scalar field with a negative sign of the kinetic term @xcite , or the combination of quintessence and phantom in a unified model named quintom @xcite . in field
dark energy models one can find potential - independent solutions @xcite , but in general the cosmological evolution depends significantly on the potential choice . in the quintessence case , one well studied potential is the inverse - power law one @xcite .
this potential exhibits `` tracker '' behavior , that is for a large class of initial conditions the cosmological evolution converges to a common solution at late times @xcite .
furthermore , in such models the quintessence energy density remains small at early and intermediate times and thus the known cosmological epochs are not affected .
finally , a theoretical justification of power - law potentials can arise through supersymmetric considerations @xcite . the cosmological evolution in such quintessence models at early times ,
is dominated by the matter - fluid and one can obtain the tracker solutions analytically @xcite . at intermediate times
the quintessence field is still small , but non - negligible , and one has to rely on perturbative analytical expressions @xcite .
however , such an extension to low redshift is very useful , since observations like supernovae ia , wmap and sdss ones , are related to this period @xcite . in this work
we are interested in investigating the phantom scenario in power - law potentials , both at high and low redshifts , and provide the tracker solutions and the perturbative analytical expressions respectively . due to the `` inverse '' kinetic behavior of phantoms in potential slopes , we expect that we have to consider normal power - law potentials instead of inverse power - law ones . indeed , starting with an arbitrary power law , we find that physically meaningful cosmological evolution excludes negative exponents
. the plan of the work is as follows : in section [ zeroord ] we construct phantom models in power - law potentials and we present the matter - dominated solutions . in section [ firstord ]
we derive the cosmological solutions at intermediate times , that is when the phantom dark energy is non - negligible , but still sub - dominant comparing to the matter content of the universe . in section [ analnum ]
we compare our analytical expressions with the exact numerical evolution of the phantom scenario .
finally , in section [ discuss ] we provide some applications of the model discussing its cosmological consequences , and we summarize our results .
let us construct the simple phantom cosmological scenario in a flat universe .
the action of a universe constituted of a phantom field @xmath3 is @xcite : @xmath4 , \label{actionquint}\end{aligned}\ ] ] where we have set @xmath5 .
@xmath6 is the phantom field potential and the term @xmath7 accounts for the ( dark ) matter content of the universe , considered as dust . finally ,
although we could straightforwardly include baryonic matter and radiation in the model , for simplicity reasons we neglect them since we are interested in @xmath8 era .
the friedmann equations are @xcite : @xmath9 @xmath10 while the evolution equation for the phantom field is : @xmath11 where @xmath12 is the hubble parameter and @xmath13 the scale factor . in these expressions
, @xmath14 is the dark matter density , and @xmath15 and @xmath16 are respectively the density and pressure of the phantom field given by : @xmath17 therefore , we can equivalently write the evolution equation in field terms as : @xmath18 the system of equations closes by considering the evolution of the matter density , which in the case of dust reads simply @xmath19 , with @xmath20 its value at present .
finally , the dark energy equation - of - state parameter is given by @xmath21 as mentioned in the introduction , in the present work we study phantom evolution in power - law potentials .
that is we consider : @xmath22 where @xmath23 is a constant with units of @xmath24 .
note that we prefer to insert the minus sign in the exponent in order to coincide with the usual notation of the literature .
a priori we do not restrict the value of @xmath25 .
however , as we are going to see , energy positivity will force @xmath25 to be negative and bounded , that is potential ( [ pot ] ) will be a normal ( and not inverse ) power law .
it proves convenient to express the aforementioned cosmological system using the scale factor @xmath26 as the independent variable , since it is straightforwardly related to the redshift @xmath27 which is used in observations .
following @xcite we define @xmath28 and thus the first friedmann equation gives @xmath29 where @xmath30 stands for @xmath6 .
thus , one can simply write : @xmath31 comparing to the quintessence case , and apart from the obvious expectation that @xmath32 is now negative , we observe , that it is also bounded .
in particular , requiring phantom total energy positivity ( @xmath33 ) we obtain @xmath34 . finally , using expressions ( [ rhox])-([hx ] )
, the field evolution equation ( [ phiddot ] ) becomes @xmath35 with prime denoting the derivative with respect to @xmath26 .
this equation is exact and accounts for the complete dynamics of the phantom scenario . in this section
we desire to present the tracker solutions during the matter - dominated era .
thus , we consider @xmath36 , or equivalently @xmath37 . in this case equation ( [ eomx ] ) is simplified as : @xmath38 with @xmath39 .
the zero subscript in parentheses denotes just this zeroth - order solution in terms of @xmath32 , or equivalently in terms of the phantom energy density , and must not be confused later on with the subscript @xmath40 without parentheses which stands for the present value of a quantity .
equation ( [ eomxzero ] ) can be easily solved analytically in the case of power - law potentials .
for @xmath41 we acquire : @xmath42 where @xmath43 and @xmath44 are constants determined by initial conditions . for @xmath45 and @xmath46 ( which is the trivial case of a constant potential ) ,
the general solution of ( [ eomxzero ] ) is @xmath47 where the function @xmath48 is related to the potential parameters through @xmath49^{\frac{1}{2+\alpha}}. \label{calpha}\ ] ] in ( [ solphizero ] ) we have kept only the solution part that remains small ( together with its derivative ) for small @xmath26 s , in order to be consistent with the matter - dominated approximation ( @xmath36 ) .
that is why initial - condition dependent constants are absent ( which is just the central idea of tracker solutions ) .
in addition , it is easy to see that this solution remains regular for @xmath50 .
the zeroth - order solution for @xmath51 can be calculated from ( [ rhox ] ) under @xmath52 , that is @xmath53 , where @xmath54 .
the result is : @xmath55 from this expression we educe the physically interesting result that @xmath56 in order for @xmath57 to be positive ( additionally in this case the @xmath48-definition in ( [ calpha ] ) is unambiguous ) .
the restriction of @xmath25 to negative values , that is the consideration of normal power - law potentials instead of inverse power - law ones ( contrary to quintessence case ) , was expected due to the `` inverse '' kinetic behavior of phantoms in potential slopes .
the additional lower bound of @xmath25 values , which breaks this form of symmetry between phantom and quintessence models , arises from the extra and necessary assumption of positivity of the total phantom energy , which is not present in quintessence case @xcite where one can always obtain it ( the potentials can always be shifted to positive values in cosmology ) .
in other words , if the potential is too steep ( large @xmath58 ) , then at early times ( where the field value @xmath59 , i.e it is close to the potential minimum ) the phantom potential energy will be very small ( @xmath60 ) and thus unable to make the total energy positive . on the other hand ,
if the potential is smooth ( small @xmath58 ) then positivity is obtained .
quantitatively , @xmath41 separates the two regions , and indeed one can see that the energy density arising form ( [ solalpha2 ] ) is negative unless @xmath61 , in which case it is exactly zero . but @xmath61 is implied by the requirement @xmath62 at small @xmath26 . therefore , for the limiting case @xmath41 , the matter - dominating condition leads to @xmath63 , and thus this exponent value bounds the two aforementioned regions .
the corresponding zeroth - order solution for @xmath16 is @xmath64 , which leads to : @xmath65 therefore , for the zeroth - order solution of the equation - of - state parameter we obtain : @xmath66 that is @xmath67 is constant .
note that @xmath56 leads to @xmath68 , which is a self - consistency test of our calculations since it is the basic feature of phantom cosmology .
finally , we can calculate the zeroth - order behavior of the phantom density parameter as @xmath69 , with @xmath19 , acquiring @xmath70^{-1}\label{om0tilde}.\ ] ] expressions ( [ solphizero ] ) , ( [ rhophizero ] ) , ( [ wzerosol ] ) and ( [ om0tilde ] ) are the tracker solutions for phantom cosmology in power - law potentials .
equivalently they can be expressed as a function of time , considering @xmath71 since we are in the matter - dominated era , as : @xmath72 with the equation - of - state parameter given by ( [ wzerosol ] ) .
these expressions provide an excellent approximation to the behavior of the phantom field as long as @xmath73 .
as cosmological evolution continues , the phantom field increases and phantom energy density becomes non - negligible , although still dominated by the dark matter one .
thus , we expect that the various quantities will progressively start diverging from the expressions obtained above . in this section
we are interested in studying analytically this intermediate evolution stage , and thus we perform a first - order perturbation to the zeroth order solutions of section [ zeroord ] . the corresponding quantities are denoting by the subscript @xmath74 in parentheses , and the total ones by tilde , i.e : @xmath75 this perturbation is equivalent with keeping terms up to first order in @xmath32 ( i.e keeping only @xmath76 ) and in @xmath77 , in the exact evolution equation ( [ eomx ] ) .
in addition , one can also expand the potential as @xmath78 .
therefore , inserting the expansions ( [ firstorder ] ) in ( [ eomx ] ) , we obtain @xmath79 + { 3a\phi_{(0)}'v\over 2 \rho_{m}}+ \nonumber \\ + { 3\rho_{\phi ( 0)}\over { ( \rho_{m})}^{2 } } { dv\over d\phi}- { 3 \over \rho_{m}}{d^2 v \over d\phi^2}\phi_{(1 ) } = 0 , \label{eomxfristord } \end{aligned}\ ] ] where @xmath54 . thus ,
for the power - law potential ( [ pot ] ) we acquire @xmath80 + { 27 \kappa a^{3 } \phi_{(0)}^{1-\alpha}\over 2(2+\alpha)\rho_{m0}}- \nonumber \\ & - & { 3\alpha(1+\alpha)\kappa a^{3 } \phi_{(0)}^{-2-\alpha}\phi_{(1)}\over \rho_{m0 } } = 0 .
\label{eomxfristord2}\end{aligned}\ ] ] the general solution of ( [ eomxfristord2 ] ) , for @xmath56 , is @xmath81 where we have kept only the part that remains small ( together with its derivative ) for small @xmath26 s , in order to be consistent with the matter - dominated approximation .
the perturbation of @xmath51 can be calculated from ( [ rhox ] ) keeping the corresponding terms , thus : @xmath82 where @xmath83 as it arises from @xmath84 .
therefore , using also ( [ solphizero ] ) , ( [ rhophizero ] ) and ( [ phifirstorder ] ) in order to express the result in terms of @xmath85 and @xmath86 , we acquire @xmath87 similarly , the perturbation of @xmath88 can arise form ( [ wx ] ) as @xmath89 since @xmath90 . therefore , the final result writes @xmath91 expressions ( [ rhophifirstordersolna ] ) and ( [ wfirstorder ] ) coincide with the corresponding one for quintessence in inverse power - law potentials @xcite , however @xmath92 and @xmath93 has a sign difference and in addition @xmath25 is negative and bounded . thus , @xmath86 is negative and , contrary to quintessence , @xmath94 is negative while @xmath95 is positive . in conclusion ,
the total solutions for intermediate times , that is at low redshifts , read : @xmath96,\ \ \label{phitilde}\end{aligned}\ ] ] @xmath97,\ \\end{aligned}\ ] ] and @xmath98,\end{aligned}\ ] ] where @xmath48 is given by ( [ calpha ] ) .
finally , @xmath99 can be easily calculated as @xmath100 , giving lastly , note that since we have these quantities as a function of the scale factor , it is straightforward to express them as a function of the redshift @xmath27 through @xmath101 , with @xmath102 the present value .
in the previous two sections we extracted analytical expressions for phantom evolution in power - law potentials , when the universe is dominated by the dark matter sector , and we argued that these solutions are valid at early and intermediate times , that is at both high and low redshift . in order to test the precision of our formulae , we evolve numerically the exact cosmological system calculating the exact quantities @xmath103 , @xmath104 , @xmath105 , and then we examine their divergence from the zeroth and first order expressions by investigating the corresponding ratios .
we choose initial conditions consistent with initial matter domination ( if this requirement is fulfilled then the results do not depend on the specific initial conditions ) , and we fix @xmath23 in order to acquire @xmath106 and @xmath107 at present . in fig .
[ fig1phi ] we present the ratios of the zeroth ( tracker ) and first order field solutions to the exact numerical value , that is @xmath108 and @xmath109 , as a function of the redshift .
the calculations have been performed for three potential cases , namely @xmath110 , @xmath111 and @xmath112 . as we observe , @xmath108 is very close to @xmath74 for @xmath1 , that is the tracker solution is a very good approximation at this early evolution stage . for @xmath113
the zeroth order solution is not a good approximation ( with error @xmath114 ) but the first order one remains within @xmath115 accuracy ( with even better results for larger @xmath116 values ) .
however , after that stage , the phantom cosmological effects become significant and our approximation breaks down rapidly .
similarly , in fig . [ fig2om ] we present @xmath117 and @xmath118 versus z. the divergence of the zeroth order solution from the exact one starts at @xmath119 .
however , as we observe , the first order solution is very accurate ( with error less than @xmath0 ) up to @xmath120 .
after that point , the first order solution starts diverging rapidly from the exact one and our approximation is not valid .
the fact that at @xmath121 in some cases ( large @xmath116 ) the zeroth order solution seems to be closer to the exact evolution comparing to the first order one , is a result of the signs of the corresponding terms ( namely positive @xmath85 and negative @xmath86 ) , but it has no meaning since our approximations are not valid in that regime . finally , in fig .
[ fig3w ] we depict @xmath122 and @xmath123 as a function of the redshift . as we see , the tracker solution is very accurate up to @xmath124 , while the first order one agrees with the exact evolution within @xmath0 up to @xmath125 . in summary , we observe that the tracker ( zeroth order ) solution is accurate within an error of @xmath0 at early times and up to @xmath119 . at intermediate times
we have to use the first order analytical solution , which agrees with the exact cosmological evolution within an error of @xmath0 up to @xmath2 .
after that point , the phantom dark energy sector enhances , it dominates the cosmological evolution , and our approximation breaks down rapidly .
since we have tested the accuracy of our analytical expressions and we have determined their applicability region , one can use them to describe an arbitrary phantom evolution in power - law potentials up to @xmath126 . here
we will present some additional applications concerning observable quantities .
having extracted expressions for @xmath105 and @xmath104 which are valid up to low redshifts , we can derive the corresponding @xmath127 relation .
indeed , using ( [ rhophizero ] ) , ( [ wtilde ] ) and ( [ omtilde ] ) we result to @xmath128 thus , at early times , where phantom energy density ( that is @xmath129 ) is negligible , @xmath88 is equal to a constant value , namely @xmath67 . at intermediate times
it increases linearly with @xmath129 . comparing with the corresponding result for quintessence in inverse power - law potentials
@xcite there are two differences .
firstly , @xmath127 is an increasing function instead of a decreasing one .
secondly , since @xmath25 lies now in the interval @xmath130 , the slope is not a slowly - varying function of @xmath25 but it ranges from @xmath131 at @xmath132 to @xmath133 at @xmath134 .
finally , one can examine the accuracy of expression ( [ womega ] ) by comparing it to the exact evolution .
indeed , we find that it is very satisfactory ( with an error of less than @xmath0 ) up to @xmath2 ( or equivalently up to @xmath135 ) , and then the error grows reaching to @xmath136 at @xmath121 . additionally , and similarly to @xcite , we can use expression ( [ wtilde ] ) for @xmath137 in order to acquire a parametrization for @xmath138 for redshifts up to @xmath2 , where our approximation is still valid .
indeed we find the same result with @xcite , namely : @xmath139 with the obvious difference that now the undetermined parameter @xmath140 , which is just the present value of @xmath88 , is smaller than @xmath141 .
this fit for phantom evolution in power - law potentials proves to be satisfactory within an error of @xmath142 .
finally , one can examine ( [ wfit ] ) in parallel with the phenomenological parametrization @xmath143 of @xcite , where he finds that the two expressions have a comparable level of accuracy @xcite .
in this work we have investigated phantom models with power - law potentials .
we have extracted the tracker solutions under the assumption of matter domination , that is the general solution and common behavior of all such models at early times , that is at high redshifts . due to the nature of phantom fields , which present an `` inverse '' kinetic behavior in potential slopes , we find that total energy positivity requires normal power - law potentials instead of inverse power - law ones ( contrary to quintessence case ) , with the potential exponent being bounded by the quadratic form .
furthermore , we have extracted the general cosmological solution at intermediate times , that is at low redshifts , which is the period during the transition from matter to dark - energy domination .
such a solution can be very useful in dark energy observations , since these ( supernovae ia , wmap and sdss ones ) are related to this cosmological period @xcite .
the comparison of the analytic expressions with the exact evolution arising from numerical elaboration , shows that the tracker solution agrees with the later within @xmath0 for redshifts @xmath1 , while the first order solution is accurate within @xmath0 up to @xmath2 .
in the aforementioned analysis we have not used observations in order to further restrict the potential form , desiring to remain general .
however , from the @xmath105 expression derived in ( [ wtilde ] ) we conclude that smaller values of @xmath116 are favored if we want to acquire a @xmath88 at @xmath126 which is not too negative , consistently with observations within 2@xmath144 .
note also that smaller @xmath116-values facilitate and enhance the total energy positivity .
finally , we have to make a comment about the quantum behavior of the examined model . as it is known , the discussion about the construction of quantum field theory of phantoms is still open in the literature .
for instance in @xcite the authors reveal the causality and stability problems and the possible spontaneous breakdown of the vacuum into phantoms and conventional particles . on the other hand ,
there have also been serious attempts in overcoming these difficulties and construct a phantom theory consistent with the basic requirements of quantum field theory @xcite , with the phantom fields arising as an effective description .
the present analysis is just a first approach on phantom fields in power - law potentials . definitely , the subject of quantization of such models is open and needs further investigation .
+ a.g .
[ supernova search team collaboration ] , astron .
j. * 116 * , 1009 ( 1998 ) ; s. perlmutter _
et al . _ [ supernova cosmology project collaboration ] , astrophys .
j. * 517 * , 565 ( 1999 ) ; d. n. spergel _ et al .
_ , astrophys
. j. suppl .
* 148 * , 175 ( 2003 ) ; s. w. allen , _ et al . _ ,
mon . not .
. soc . * 353 * , 457 ( 2004 ) .
p. bintruy , c. deffayet , d. langlois , nucl .
b * 565 * , 269 ( 2000 ) ; g.r .
dvali , g. gabadadze , m. porrati , phys .
b * 485 * , 208 ( 2000 ) ; s. capozziello , int . j. mod .
d * 11 * , 483 ( 2002 ) ; s.nojiri and s. d. odintsov , phys .
d * 68 * , 123512 ( 2003 ) ; p.
s. apostolopoulos , n. brouzakis , e. n. saridakis and n. tetradis , phys .
d * 72 * , 044013 ( 2005 ) ; s.nojiri and s. d. odintsov , int . j. geom .
phys . * 4 * , 115 ( 2007 ) ; f. k. diakonos and e. n. saridakis , arxiv:0708.3143 [ hep - th ] .
p. j. e. peebles and b. ratra , , l17 ( 1988 ) ; b. ratra and p. j. e. peebles , phys .
d * 37 * , 3406 ( 1988 ) . c. wetterich , nucl .
b * 302 * , 668 ( 1988 ) ; m. s. turner and m. white , phys .
d * 56 * , 4439 ( 1997 ) ; r. r. caldwell , r. dave and p. j. steinhardt , phys .
lett . * 80 * , 1582 ( 1998 ) ; a. r. liddle and r. j. scherrer , phys .
d * 59 * , 023509 ( 1999 ) ; i. zlatev , l. m. wang and p. j. steinhardt , phys.rev .
lett . * 82 * , 896 ( 1999 ) ; z. k. guo , n. ohta and y. z. zhang , mod .
a * 22 * , 883 ( 2007 ) ; o. hrycyna and m. szydlowski , phys . rev .
d * 76 * , 123510 ( 2007 ) .
r. r. caldwell , phys .
b * 545 * , 23 ( 2002 ) ; r. r. caldwell , m. kamionkowski and n. n. weinberg , phys . rev .
lett . * 91 * , 071301 ( 2003 ) ; s. nojiri and s. d. odintsov , phys .
b * 562 * , 147 ( 2003 ) ; v. k. onemli and r. p. woodard , phys .
d * 70 * , 107301 ( 2004 ) ; p. f. gonzalez - diaz and c. l. siguenza , nucl .
b * 697 * , 363 ( 2004 ) ; s. nojiri and s. d. odintsov , phys .
d * 72 * , 023003 ( 2005 ) ; h. garcia - compean , g. garcia - jimenez , o. obregon , and c. ramirez , jcap 0807 , 016 ( 2008 ) ; e. n. saridakis , arxiv:0811.1333 [ hep - th ] ; m. szydlowski and o. hrycyna , jcap * 0901 * , 039 ( 2009 ) ; x. m. chen , y. gong and e. n. saridakis , arxiv:0812.1117 [ gr - qc ] . b. feng , x. l. wang and x. m. zhang , phys .
b * 607 * , 35 ( 2005 ) ; z. k. guo , _ et al .
b * 608 * , 177 ( 2005 ) ; m .- z li , b. feng , x .-
m zhang , jcap , * 0512 * , 002 ( 2005 ) ; y. f. cai , h. li , y. s. piao and x. m. zhang , phys .
b * 646 * , 141 ( 2007 ) ; w. zhao and y. zhang , phys . rev .
d * 73 * , 123509 ( 2006 ) ; b. feng , m. li , y .- s .
piao and x. zhang , phys .
b * 634 * , 101 ( 2006 ) ; m. r. setare and e. n. saridakis , phys .
b * 668 * , 177 ( 2008 ) ; m. r. setare and e. n. saridakis , jcap * 0809 * , 026 ( 2008 ) ; m. r. setare and e. n. saridakis , phys . lett .
b * 671 * , 331 ( 2009 ) .
r. j. scherrer and a. a. sen , phys .
d * 77 * , 083515 ( 2008 ) ; r. j. scherrer and a. a. sen , phys .
d * 78 * , 067303 ( 2008 ) ; m. r. setare and e. n. saridakis , arxiv:0810.4775 [ astro - ph ] . j. p. kneller and l. e. strigari , phys .
d * 68 * , 083517 ( 2003 ) ; l. r. w. abramo and f. finelli , phys .
b * 575 * , 165 ( 2003 ) ; x. zhang , mod .
a * 20 * , 2575 ( 2005 ) ; m. yashar , b. bozek , a. abrahamse , a. j. albrecht and m. barnard , arxiv:0811.2253 [ astro - ph ] . c. r. watson and r. j. scherrer , phys .
d * 68 * , 123524 ( 2003 ) .
i. zlatev , l. wang , and p. j. steinhardt , phys .
* 82 * , 896 ( 1999 ) ; a. r. liddle and r. j. scherrer , phys .
d * 59 * , 023509 ( 1999 ) ; p. j. steinhardt , l. wang , and i. zlatev , phys .
d * 59 * , 123504 ( 1999 ) . | we investigate phantom models with power - law potentials and we extract the early - time , `` tracker '' , solutions under the assumption of matter domination . contrary to quintessence case , we find that energy positivity requires normal power - law potentials instead of inverse power - law ones , with the potential exponent being bounded by the quadratic form .
in addition , we analytically present the general cosmological solution at intermediate times , that is at low redshifts , which is the period of the transition from matter to dark - energy domination .
the comparison with the exact evolution , arising from numerical elaboration , shows that the tracker solution agrees with the later within @xmath0 for redshifts @xmath1 , while the intermediate solution is accurate within @xmath0 up to @xmath2 . |
wolf - rayet ( wr ) stars are massive , luminous , hot stars which have blown away the hydrogen envelope , and are considered to be immediate precursors of some kinds of supernovae .
carbon - rich , late - type subclass of wr stars ( wcl ) are known as one of the most efficiently dust - productive stellar environments ( for recent reviews , see 1995,1997ab ) .
the dust production in these stars not only highlights the problems of the dust formation mechanism in a very hot environment @xcite , but also plays an important role in the chemical enrichment in galaxies ( @xcite ; @xcite ) .
in addition to the strong infrared emission in wcl stars , which is considered to arise from persistent dust shells , there has been emerging evidence of episodic dust formation in these stars . @xcite
reported transient optical fadings in several wcl stars , attributable to temporary condensations of dust clouds .
@xcite reported remarkable weakening of the wr - type spectral feature , together with a possible 1.1 magnitude photometric fading , in a wr+ob binary wr 104 , which was interpreted as a temporary obscuration of the inner wolf - rayet wind by a dust cloud . on the other hand
, recent interferometric imaging of wr 104 by @xcite revealed the amazing dusty pinwheel nebula " . this extended structure which rotates with a period of 220@xmath130
d [ @xcite further reported a refined the period to be 243.5@xmath13 d ] is believed to be formed by the colliding winds of the wr star and companion ob - star .
the relation between occasionally reported obscuration episodes and the persistent dusty pinwheel structure , however , remained a mystery ( 1999,2002 ) . in 2001 april , one of the authors ( kh ) serendipitously discovered a new variable star named hadv82 @xcite , which was subsequently identified with wr 104 .
this led to a discovery of dramatic photometric variation in wr 104 , with an unprecedented amplitude and frequency of fadings among all wr stars .
the variation has a quasi - period close to the reported binary period , which provides an essential clue to understanding the relation between the formation of dust and the role of binarity .
a total of 176 observations were done between 1994 may 12 and 2001 april 26 , with twin patrol cameras equipped with d=10 cm f/4.0 telephoto lens and unfiltered t - max 400 emulsions , located at two sites in toyohashi , aichi ( kh ) and saku , nagano ( kt ) .
the passband of observations covers the range of 400650 nm .
photographic photometry was done using neighboring comparison stars whose @xmath3-magnitudes were calibrated by t. watanabe .
the overall uncertainty of the calibration and individual photometric estimates is 0.20.3 mag , which will not affect the following analysis .
the resultant light curve is presented in figure [ fig : lc ] .
the star showed an overall range of variability between 11.8 and fainter than 14.5 magnitudes , making the full amplitude greater than 2.7 magnitudes .
the raw data are available at @xmath4 ftp://vsnet.kusastro.kyoto-u.ac.jp/pub/vsnet/wr/wr104/wr104obs.jd @xmath5 .
such a large variation in visible light far exceeds previously known variations of wr - type stars @xcite , which only occasionally show temporary fadings having depths less than 1 magnitude .
faint phases of wr 104 typically last an order of months , which is strikingly longer than any known transient episodes of wr - type variable stars @xcite , which usually last days to weeks .
the most remarkable feature is the existence of a quasi - period of 200400 d , superimposed on a long - term trend , which may be attributed to slow changes of optical depth of the line - of - sight extinction .
the result of period analysis using the phase dispersion minimization technique @xcite is presented in figure [ fig : pdm ] .
the strongest period in the range of 50500 d is 241 d , which is in almost perfect agreement with the suggested orbital period of 243.5@xmath13 d @xcite .
figure [ fig : phase ] shows a phase - averaged light curve with this period .
the mean orbital light curve is characterized by a rather sharp minimum and a broader maximum .
the previously known most striking evidence of variability in wr 104 is the temporary remarkable weakening of the spectroscopic feature of the wr - type star @xcite .
@xcite suggested that occultation by a dust cloud condensation , analogous to r crb stars , is responsible for the phenomenon .
the coincidence of the the remarkable weakening of the wr - type spectral feature and the optical fading supports the idea that the wr - type star was almost entirely obscured at the time of the observation by @xcite .
the remarkable agreement of the dominant period in optical variability with the orbital period supports that at least a considerable fraction of varying dust obscuration is physically related to the binary motion .
we mainly discuss on implications of the remarkable coincidence between the period of large - amplitude optical variation and the binary period .
first , the reported low binary inclination of 11@xmath17@xmath2 @xcite makes unlikely that the tail of the dust spiral is directly responsible for the variation .
a possible explanation is the periodic enhancement of dust production in the wr - star wind at the passage of the companion star though an elliptical orbit .
this effect is most pronounced in an episodic dust producer , wr 140 @xcite and presumably wr 137 @xcite .
the evidence for this scenario in wr 104 is less convincing , in its little infrared variability @xcite , and in its continuous appearance of the dust spiral ( 1999,2002 ) , and the lack of evidence of a large orbital eccentricity @xcite , all of which were considered against episodic dust production around periastron .
further full - orbit interferometric observation , together with contemporaneous visual and infrared photometry will be key information in testing the hypothesis .
the other possibility is that the varying obscuration is of geometrical origin , e.g. varying extinction as a consequence of rotating binary seen through a gradient of absorption near the line of sight .
this explanation is similar to an idea to explain the unique variability of the binary central star of the planetary nebula , ngc 2346 ( @xcite ; @xcite ) .
this explanation would require an extremely strong gradient or a sharply defined dense obscuring body in the line of sight .
the observed amplitude ( more than 2.7 magnitudes ) could only be explained if the wr - type component emits about 90% of the visual light , and the star is almost completely eclipsed by this obscuring structure .
however , this assumption may not be consistent with its weaker wr - type spectral feature even in the high state than in other wc9 stars @xcite .
such apparent discrepancies were also observed in wr 137 @xcite , which may be a rather common phenomenon seen in wcl+ob wind collision binaries . considering the total line - of - sight absorption of @xmath6=6.5 magnitude @xcite ,
about 4.5 mag of which ( after subtracting the interstellar absorption ) is likely to be attributed to past mass - loss events of the progenitor , rather than the present dust formation ( @xcite , @xcite ) , such a large fraction ( 2.7/4.5 ) of variable line - of - sight extinction seems to be difficult to reconcile with the past , presumably more spherical mass - loss episodes .
the present discovery of strongly variable extinction may alternately suggest that the highly structured extinction can be being formed via an ejection of dust in the direction of the binary rotation axis . such a possibility may be tested by a future development of full three - dimensional treatment of dust formation and ejection in a colliding - wind binary .
another viable explanation is that the three - dimensional structure of the shock front , where the compressed gas is considered to effectively form dust grains @xcite , itself is the obscuring body .
depending on the geometry , the dusty shock front near the conjunction phase of the binary can completely obscure the inner wr - star wind , as well as the ob star , both are required to explain the spectral variation @xcite and the observed amplitude of optical fading .
this explanation would require a higher inclination angle than was reported , or a sufficient vertical structure to obscure the stellar component viewed even nearly pole - on .
detailed high - resolution , full - orbit observation is again indispensable to discriminate the possibilities .
the present discovery of quasi - periodic , highly variable visual obscuration in wr 104 proposes a new class and mechanism of light variability in wr - type stars . with the advent of sub - milliarcsecond optical and infrared interferometry , and the coming era of atacama large millimeter array ( alma ) , wr 104
would provide a powerful tool in geometrically resolving the heart of dust production in the most efficiently dust - productive stellar environments .
haseda , k. 2001 , vsnet - alert circulars , no .
5826 , 5832 , + ( http://www.kusastro.kyoto-u.ac.jp/vsnet/mail/alert5000/msg00826.html and + http://www.kusastro.kyoto-u.ac.jp/vsnet/mail/alert5000/msg00832.html ) | we discovered that the wolf - rayet ( wr)+ob star binary , wr 104 , renowned for its associated dusty pinwheel nebula " recently spatially resolved with infrared interferometry , exhibits strong quasi - periodic optical variations with a full amplitude of 2.7 mag .
such a large - amplitude , continuous variation has been unprecedented in a wr star . the optical quasi - period ( @xmath0241 d )
is in almost perfect agreement with the interferometric period ( 243.5@xmath13 d ) .
the remarkable agreement of the dominant period in optical variability with the orbital period supports that the strongly varying dust obscuration is physically related to the binary motion , rather than sporadic dust - forming episodes . considering the low orbital inclination ( 11@xmath17@xmath2 ) and the nearly circular orbit inferred from the interferometric observations ,
the strongly variable line - of - sight extinction suggests that the highly structured extinction can be being formed via an ejection of dust in the direction of the binary rotation axis .
another viable explanation is that the three - dimensional structure of the shock front , itself is the obscuring body .
depending on the geometry , the dusty shock front near the conjunction phase of the binary can completely obscure the inner wr - star wind and the ob star , which can explain the amplitude of optical fading and the past observation of remarkable spectral variation .
keywords : stars : wolf - rayet stars : variables stars : winds , outflows stars : individual ( wr 104 ) ( 140mm,70mm)fig1.eps |
the approximation of local thermodynamic equilibrium ( lte ) is commonly used in the analysis of stellar spectra .
it simplifies the computation of atomic and molecular level populations immensely by postulating saha - boltzmann equilibria , thus defining the state of the gas as a function of only temperature , pressure , and chemical composition .
the solution of the equation of radiative transfer for the frequencies of interest is then straightforward .
the implied assumption that the radiation field does not influence the properties of the gas is , however , not realistic .
thus we must question the validity of lte and investigate the nlte problem of spectral - line formation .
this will here for most of the time be taken to mean the solution of the coupled equations of statistical equilibrium and radiative transfer for a trace species in a stellar photosphere .
nlte effects are then simply errors caused by the lte approximation . when discussing the nlte effects on a spectral line , it is often clarifying to separate between effects on the line opacity and on the line source function .
the line - opacity effect is described with the nlte departure coefficient of the atomic population of the lower energy level involved in the transition : @xmath0 .
the departures from lte of the line source function is described by its ratio to the local planckian , @xmath1 . in the following , nlte effects on lines used for oxygen - abundance determinations in unevolved or moderately evolved stars of approximately solar effective temperatures will be discussed .
the discussion focuses only on departures from lte , thus disregarding for a moment all other relevant uncertainties like missing background opacities , observational errors , stellar parameters , effects of granulation , line - damping treatment , etc .
the forbidden lines of oxygen , at 630 nm and 636 nm have not been claimed to be subject to any nlte effects in cool stars .
a typical nlte calculation shows them to be exceedingly close to lte both in line opacity and line source function .
the lte value of the line opacity is due to the fact that the lower levels of these transitions belong to the ground term of neutral oxygen .
virtually all free oxygen atoms will be in that state because of the lack of excited levels of low energy and the high ionisation potential .
( the amount of oxygen tied up in molecules is not enough to allow any departures from lte in the molecular equilibrium to have an impact . ) thus , exceedingly large departures from lte would have to be present to influence the line opacities significantly .
note that the ionisation equilibrium of oxygen is coupled to that of hydrogen via charge - exchange collisions due to the close correspondence of the ionisation potentials of these two atoms .
this means that if nlte effects would be significant , the problem would be a hydrogen problem rather than an oxygen one .
no departure of the line source function from the local planckian is expected or seen in any computational results .
this is because the lines are forbidden and weak with collisional rates dominating very much over radiative rates , and the upper level involved is collisionally excited .
the neutral oxygen triplet lines at 777 nm called just the triplet hereafter for brevity are well - known for being affected by nlte effects , especially in early stars . for solar - type stars , the picture is a bit more unclear since the quantitative predictions have had problems when confronted with solar observations .
there are of course other permitted optical lines that have been used for abundance analysis .
they are , however , mostly weaker and thus less useful than the triplet at lower abundances .
the nlte effects on those lines are similar in mechanism though lesser in magnitude than those affecting the triplet .
figure [ fig : depcoeff ] shows the theoretical behaviour of the strongest triplet line in a solar photospheric model . ( computed without any contribution of inelastic collisions with neutral hydrogen atoms a problematic process discussed below . )
we see that the line opacity stays close to its lte value because the departure coefficient for the line s lower level ( @xmath2 ) @xmath3 .
the line source function ( @xmath4 ) shows a significant departure from lte as it dips below the planckian , thus producing a line that is stronger than it would be in lte .
the line source function behaves like that of the text - book two - level atom without continuum ( e.g. * ? ? ?
11 ) , @xmath5 where @xmath6 measures the destruction probability of line photons by collisional deexcitation .
it is the photon losses that cause @xmath4 to fall below planckian and scattering in the line makes this effect go further inwards than the monochromatic optical depth in the line would imply . the effect grows stronger with increased abundance as the line strength increases .
but it does not generally go to zero as the abundance decreases since @xmath7 in the infrared continuum . by constructing curves of growth from the lte and the nlte equivalent widths
, one can compute nlte abundance corrections , @xmath8 , that can be added to abundances derived under the lte assumption .
these will _ always be negative _ , meaning that an lte analysis will always overestimate the oxygen abundance .
figure [ fig : abcorr ] shows these corrections computed from the results of fig . [ fig : depcoeff ] . for solar - type stars
, @xmath9 varies from 0 up to @xmath10 dex ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , getting more significant with increased effective temperature and decreased surface gravity .
no strong monotonous dependence on metallicity has been predicted that would be able to seriously change the slopes in [ o / fe ] plots or remove the discrepancies between the trends from different oxygen - abundance indicators .
the two - level - atom behaviour is good for us because it makes it easier to analyse and understand the results physically .
it also assures us that it is the processes in the line transition itself that are important high - lying levels , photoionising radiation fields , and other things problematic to model do not matter .
this also allows simplification in , for example , work on 3d radiative transfer @xcite .
note that the effect has nothing to do with the lines being of high excitation energy or that their lower level is metastable .
this is just the natural behaviour of reasonably strong lines in the infrared with two - level - atom like behaviour and lte - like opacity .
there is a limit to this two - level regime , however .
note that the departure coefficient of the lower level in fig .
[ fig : depcoeff ] is not exactly equal to one .
if the effective temperature is increased , we will eventually get a significant departure from lte of the line opacity .
the result will then depend on the rest of the atomic model and the interpretation of the results will be less straightforward .
an important ingredient is probably the recombination flow from higher levels @xcite , but this transition to more complicate behaviour has not been analysed well yet .
@xcite investigated the centre - to - limb behaviour of the triplet in the sun and found that this was not consistent with the triplet being formed in lte in that the line strengths do not fall off as quickly close to the limb .
@xcite could reproduce the solar observations well with an nlte analysis .
but then the accepted oxygen abundance from other spectral features increased and a discrepancy was noted @xcite and persisted @xcite . @xcite found it impossible to reconcile observations with any reasonable one - dimensional photospheric model and an oxygen abundance @xmath11 , leaving granulation effects as the only possibility .
however @xcite found from 3d nlte modelling in hydrodynamic granulation simulations that the effects on line strength and @xmath12 dependence could not significantly alleviate the situation .
this solar discrepancy may not be the most dramatic problem we have in astronomy
it is probably , after all , solved by allowing an abundance below @xmath13 ( e.g. * ? ? ?
* ) but it illustrates the uncertainties surrounding spectral - line formation and their coupling to solar abundances .
indeed , the oxygen triplet lines should be among the easiest to model among solar lines affected by departures from lte .
given the two - level nature of the triplet in solar - type stars , the most problematic atomic quantity is , as usual in nlte work , collisional cross sections .
while electron collisions are always uncertain , this is even more so for collisions involving neutral hydrogen atoms the case of the oxygen triplet illustrates the effect of this uncertainty clearly .
collisional excitation by atoms in cool - star nlte work were introduced by @xcite .
they generalised the estimates of @xcite , which related to collision between like atoms , to inelastic collisions between hydrogen atoms and other species .
these results have been used by many authors .
@xcite reviewed these formalisms and suggested a reasonable improvement to them which , however , does not seem to have taken on . the formul based on drawin s work can only be regarded as order of magnitude estimates , and perhaps not even that .
many authors have tried to get better estimates by multiplying the rates of @xcite with a factor @xmath14 determined by some empirical fitting .
@xcite fitted the triplet lines to the solar spectrum assuming @xmath15 .
this resulted in large collisional rates almost producing lte .
@xcite made a multiparameter fit of solar line profiles resulting in a scaling factor of @xmath16 , thus confirming the @xcite formula .
@xcite confirmed that the solar centre - to - limb behaviour of the triplet was not consistent with lte .
@xcite used the @xmath12 dependence and profiles of solar lines to find that the hydrogen collisions were negligible ( @xmath17 ) .
the same result came from his effort to minimise the difference between the abundances derived from the individual triplet lines for a set of stars of different effective temperatures .
@xcite found that @xmath18 gave consistent abundances for different oxygen lines in rr lyr stars . among other recent work that can be mentioned , @xcite chose @xmath19 probably because the use of that factor in the work on fe in pollux by @xcite .
apparently the attempts to determine empirical cross sections are very model dependent , and it seems hard to avoid all other uncertainties related to stellar modelling and parameters notably effective temperatures to bias the procedure . if one therefore discards the extrasolar evidence and that depending on certain values of the solar oxygen abundance , it seems that the solar centre - to - limb behaviour is still a rather strong argument that hydrogen collisions are not very significant and that they definitely can not induce lte .
circumstantial evidence comes from studies of other lines .
there is now a detailed quantum mechanical calculation of the hydrogen excitation cross section of the na d lines @xcite .
this shows a cross section that is @xmath20 of the drawin - style prediction .
see also @xcite . in the time
since the kiel group pointed out inelastic collisions with hydrogen atoms as potentially very important in cool - star nlte work , many large and expensive telescopes and spectrographs have gone from visionary ideas to first light .
but our knowledge about these collisional rates so important for interpreting some of the results of these technological wonders has not advanced very much .
we do not know much about the nlte formation of molecular lines . @xcite
estimated that the vib - rot levels among the electronic ground state of diatomic molecules in the solar photosphere should be in boltzmann equilibrium but that departures from lte could be expected in the line source functions of electronic transitions . for weak ultraviolet lines ( think oh in metal - poor stars here )
this could mean that the lines get weaker for a given abundance than lte predicts and thus lte abundances are underestimates .
when the lines are strong , the effect could go the other way .
another matter is the chemical equilibrium itself , where photodissociation could play a role in causing departures from lte .
it could even be that statistical equilibrium is not reached in the few minutes that the upwelling gas in a granule spends in and above the photosphere before falling back beneath the visible surface again .
@xcite has compared observed co lines with simulations and suggests that this effect is really seen .
if so , all modelling assuming statistical equilibrium will derive too low abundances
. there could thus be nlte effects in the line radiative transfer which require collisional cross sections for modelling , nlte effects in the statistical equilibrium of molecular abundances which require modelling of photodissociation , and maybe even dynamic modelling of molecular formation is required to properly model the formation of co and oh lines .
for those in quest of accurate oxygen abundances of solar - type stars , it should be noted that : * nlte effects are utterly unimportant for the forbidden oxygen lines . *
the 777 nm triplet lines are not formed in lte .
nlte effects will not change the slopes in the [ o / fe][fe / h ] diagrams very much , but they will surely influence the error bars because of the uncertainties in the collisional cross sections . * the nlte effects on the triplet are always expected to be negative . thus lte results define an _ extreme _ on the possible range of abundances , not a conservative mean hypothesis . *
nlte effects on molecular lines should be investigated . | the nlte effects affecting oxygen - abundance determinations of solar - type stars are discussed .
lte is perfectly safe for the forbidden lines .
the permitted triplet at 777 nm is expected to show nlte effects on the order of a few tenths of a dex ( always in the sense that lte overestimates the abundance ) , but the magnitude of the effects is dependent on the still very uncertain cross sections of collisional excitation by collisions with neutral hydrogen atoms .
little is known about the nlte effects on molecular line formation .
line : formation ; sun : abundances ; stars : abundances |
the problem of time relaxation around a local minimum of a free - energy landscape is ubiquitous in chemical physics .
in fact , the landscape picture @xcite assumes a natural separation of low - temperature molecular motion sampling distinct potential energy minima , and vibration within a minimum .
the manner in which a disordered material samples its landscape as a function of temperature thus provides information on its long - time relaxation properties .
the energy landscape paradigm has been successfully applied to protein folding @xcite , the mechanical properties of glasses @xcite , and the dynamics of supercooled liquids @xcite . in this context ,
adam - gibbs formula @xcite suggests a phenomenological connection between kinetics and thermodynamics in disordered systems , that is , @xmath0 , where @xmath1 is a relaxation time , @xmath2 and @xmath3 are two phenomenological constants , and @xmath4 is a configurational entropy factor related to the number of minima of the system s multidimensional energy surface .
for instance , at low enough temperatures the system becomes stuck in a single minimum , the depth of which increases as the cooling rate decreases : this describes a glass transition . in this context , of prominent interest
is the case of relaxation between two degenerate free - energy minima separated by an ( almost ) symmetric activation barrier . in the current literature this is referred to as the kramers problem @xcite . in calculating the average transition time between two such states ,
one typically ignores the presence of other possible less stable ( more energetic ) states in the free - energy landscape @xcite .
we show that statistically rare trajectories that connect two such degenerate states only after entering another neighboring state , are responsible for an increase of the relevant mean transition time , sometimes by orders of magnitude .
the consequence is that in order to ignore the contribution of slowly meandering trajectories and keep using the results of standard kramers theory , one has to restrict the system s phase - space volume defining the free - energy stable states .
our conclusion has an immediate counterpart and , hopefully , application in the strategies of path sampling for the numerical investigation of complex systems @xcite .
for instance , an unfolded protein can explore thousands of intermediate structures ( conformations ) before reaching a long - lived ( stable ) folded conformation . the most numerically efficient approach to investigate
this process involves simulating protein folding with molecular dynamics for a relatively short time , and then analyzing the resulting trajectories to extract a coarse - grained markov state model ( msm ) .
an msm consists of an appropriate choice of long - lived clustered conformational states and the transition rates between them . to create an msm , one runs molecular dynamics simulations to determine how frequently a protein changes from one state to another , and clusters intermediate structures based on kinetic proximity ( e.g.
, how energetically easy is switching from one structure to another ) .
the transition rates are typically determined by averaging the time the protein takes to switch between any two states encoded in the msm .
due to the coarse - grained nature of the msm , a continuous trajectory connecting a pair of sampled states might well enter first the phase - space basin belonging to another state without being trapped there .
this occurrence , though unlikely , may dramatically affect the corresponding transition time .
how to correctly generate the reactive trajectories representing a specific transition of interest for the msm is an issue of ongoing research .
the contents of this paper is organized as follows . in sec .
[ bistable ] we first simulate the relaxation of an overdamped brownian particle in a 1d potential well .
we determine both numerically and analytically the mean first - passage time ( mfpt ) for the particle to reach the bottom of the well from an injection point inside it .
we show that when the injection point rests inside the well , but higher than the bottom of another adjacent well , then the rare trajectories crossing the barrier separating the two wells become dominant and , on lowering the noise level , the mfpt increases exponentially .
some of the results presented here have been independently derived in ref .
@xcite for discrete stochastic models of biological interest . in sec .
[ dominant ] we interpret this effect by distinguishing between two types of trajectories , the most probable trajectories pointing from the injection point straight down to the well bottom , and the rare trajectories overcoming the barrier into the side well .
the distribution density of the relaxation times allows a clear - cut distinction between these two types of trajectories . in sec .
[ multiwell ] we extend our analysis to the case of multiwell potentials and conclude that the mfpt inside a well is dominated by barrier - crossing anytime the particle s injection point rests above the level of the lowest lying among all adjacent wells ( sec . [ threewellnondeg ] ) .
finally , we consider the case of the hopping process between two degenerate minima of the potential and discuss how the mfpt over the barrier separating them can depend on the level of the injection point and , therefore , on the multistable structure of the entire potential ( sec .
[ threewelldeg ] ) . in sec .
[ conclusions ] we draw some concluding remarks regarding the impact of this effect on the interpretation of actual relaxation measurements .
we start introducing two categories of trajectories a 1d system may take while relaxing toward a stable state . broadly speaking ,
we distinguish between _ regular _ trajectories , the most probable and typically the shortest ones , given certain initial conditions , and a subset of _ dominant _ trajectories , which one determines with reference to the observable being measured .
the most probable transition trajectories in a 1d system has been classified by analyzing the ( local ) minima of the relevant action integrals @xcite . here , we are rather concerned with identifying the systems trajectories that most contribute to the mean value of a specific observable of interest .
a study - case is represented by the transition times @xmath5 of an overdamped brownian particle obeying the langevin equation ( le ) , @xmath6 where @xmath7 denotes the particle coordinate , @xmath8 is a confining multistable potential , and @xmath9 models a stationary , zero - mean , gaussian noise source with autocorrelation function @xmath10 the particle will be injected at a given point @xmath11 and taken out upon reaching the exit point @xmath12 . to keep our notation as simple as possible
, we place the exit point at the bottom of a potential well , termed well @xmath13 , located on the left of the injection point , i.e. , @xmath14 , see inset of fig . [ f2.1 ] .
the time length of each trajectory is the observable of interest , @xmath5 .
the average transition time @xmath15 for the particle to diffuse from @xmath11 to @xmath13 , is given by the well - known mfpt formula @xcite , @xmath16 where @xmath17 $ ] is the stationary probability density of the process ( [ 2.1 ] ) .
note that for a confining potential , @xmath18 , i.e. , @xmath19 can be treated as a reflecting boundary @xcite vs. @xmath11 from numerical integration of the le ( [ 2.1 ] ) with the asymmetric bistable potential @xmath20 ( inset ) and different noise strength , @xmath21 .
the potential minima are located at @xmath22 and @xmath23 , the barrier at @xmath24 , and the crossover threshold , defined by @xmath25 , at @xmath26 .
the dashed curves @xmath27 were obtained by performing the double integral in eq .
( [ 2.4 ] ) for the appropriate @xmath21.[f2.1 ] ] we specialize now eq . (
[ 2.3 ] ) to the case of an asymmetric bistable potential . as illustrated in the inset of fig .
[ f2.1 ] , @xmath28 locates the top of the barrier , @xmath29 , and @xmath12 and @xmath30 denote the bottom of the left , @xmath13 , and right well , @xmath31 , respectively , with @xmath32 . here and in the following
, we adopted the short - hand notation @xmath33 , @xmath34 , @xmath35 , @xmath36 , and prime for an @xmath37 derivative , @xmath38 .
the threshold @xmath39 is the point on the r.h.s .
of @xmath13 that has the same potential energy as the bottom of well @xmath31 ; for the asymmetric double - well potential of fig .
[ f2.1 ] , @xmath25 with @xmath40 .
we then estimate the mfpt ( [ 2.3 ] ) in the weak noise limit , @xmath41 , for three different ranges of the injection point , @xmath11 : _ ( i ) out - of - well , @xmath42 . _
the functions @xmath43 and @xmath44 are sharply peaked , respectively , around points @xmath12 and @xmath30 and around point @xmath28 . as a consequence , for @xmath42 the nested integrals ( [ 2.3 ] ) factorize ,
that is , @xmath45 in the limit of weak noise @xcite @xmath46 $ ] , so that the integrals ( [ 2.4 ] ) can be approximated to @xmath47 this is the well - known kramers formula , @xmath48 , for the escape time out of well @xmath31 . here , according to our notation , all escape trajectories are regular and the ensuing ( almost @xmath11 independent ) relaxation time is characterized by the slow relaxation process @xmath49 . _
( ii ) barrier well region , @xmath50 .
_ for this choice of the injection point , the first integrand ( [ 2.4 ] ) can be approximated to @xmath51 $ ] ; hence @xmath52 here we took the absolute value of @xmath53 only for the sake of generality .
this result is suggestive : although the particle was injected directly in well @xmath13 , still it takes an exponentially long average time to reach its bottom , @xmath12 .
moreover , in contrast with kramers time of eq .
( [ 2.5 ] ) , @xmath27 appears to depend on how high the injection point lies with respect to the minimum , @xmath54 , of the side - well @xmath31 . as discussed in sec .
[ dominant ] , the mfpt ( [ 2.6 ] ) is indeed dominated by the rare trajectories that cross over into well @xmath31 before being absorbed at @xmath12 . _
( iii ) bottom well region , @xmath55 .
_ as @xmath11 approaches the exit point , one can easily take the @xmath56 limit of the double integral ( [ 2.3 ] ) , thus obtaining the logarithmic law , @xmath57,\ ] ] where @xmath58 is the mascheroni s constant .
this is the short mfpt one would expect on account of the sole regular trajectories of the relaxation process .
indeed , such trajectories run straight downhill from @xmath11 subject to weak noise fluctuations , whose effect grows appreciable only close to the exit point , @xmath59 .
our analytical estimates ( [ 2.5])-([2.7 ] ) reproduce well the three different regimes of the @xmath27 curves of fig .
[ f2.1 ] , obtained by numerically computing the double integral ( [ 2.3 ] ) for very small @xmath21 values .
the crossover between the logarithmic ( [ 2.7 ] ) and the exponential branch ( [ 2.6 ] ) of @xmath27 is fairly sharp , because the exponential in eq .
( [ 2.6 ] ) abruptly vanishes for @xmath55 and @xmath60 . in passing we notice that our approximations ( [ 2.5 ] ) and ( [ 2.6 ] ) coincide ( apart from minor typographical errors ) with the first two mfpt s reported in eq .
( 33 ) of ref.@xcite for schlgl s model in the large size system limit .
our derivation is much simpler , indeed , but restricted to the case of continuous stochastic transition processes .
finally , the results of this section can be readily extended to the case when the side - well @xmath31 is deeper than the exit well , @xmath61
. only approximation ( [ 2.7 ] ) needs to be modified as the probability density , @xmath43 , in the exit well gets exponentially suppressed . as a consequence , the right hand side of eq .
( [ 2.7 ] ) must be multiplied by the additional factor @xmath62 $ ] .
this means that , since no threshold @xmath39 could be defined , the average transition time is exponentially long for any in - well injection point , namely , @xmath63 $ ] for @xmath64 .
as anticipated in the foregoing section , the results of eqs .
( [ 2.5 ] ) and ( [ 2.7 ] ) lend themselves to a simple interpretation in terms of regular trajectories . for @xmath65 the particle
is initially placed in the side - well @xmath31 , so that it , first , relaxes around the local @xmath43 maximum at @xmath30 and , then , escapes into well @xmath13 by overcoming the barrier @xmath29 ; as a consequence @xmath27 is quite insensitive to the injection point @xmath11 .
for @xmath66 the particle tends to roll downhill toward the exit point @xmath12 , corresponding to the absolute maximum of @xmath43 , with a short average transition time proportional to the logarithm of the initial displacement , @xmath67 .
, of the transient times @xmath68 obtained by numerically integrating the le ( [ 2.1 ] ) for the asymmetric bistable potential of fig .
[ f2.1 ] with @xmath69 and different @xmath11 .
the dashed curves represent the harmonic approximation @xmath70 of eq .
( [ ptharm ] ) for the two @xmath11 closest to @xmath12 , see text .
inset : semi - logarithmic plot of @xmath71 vs @xmath72 for three values of @xmath11 and @xmath69 .
the three data sets are closely fitted by the function @xmath73 in eq .
( [ 3.4 ] ) .
the dashed line with @xmath74 has been drawn to guide the eye .
[ f3.1 ] ] the transitions that start out in the barrier region @xmath75 are qualitatively different . as the injection point lies inside well @xmath13 , the trajectories oriented toward the exit point are still the most probable , or , stated otherwise , they represent the process _ regular _ trajectories , as expected .
nevertheless , the particle can diffuse from @xmath11 over the barrier into well @xmath31 with small but finite probability .
following refs .
@xcite , we can estimate the splitting probability @xmath76 for the particle to exit at @xmath13 without first reaching @xmath31 , and @xmath77 for the particle to fall into well @xmath31 before being absorbed at @xmath13 , @xmath78 @xmath79 for weak noises and @xmath11 not too close to the extrema @xmath12 and @xmath28 , the integral ( [ 3.2 ] ) can be approximated to @xcite @xmath80 vs. @xmath11 in the asymmetric three - well potentials ( a ) @xmath81 and ( b ) @xmath82 for different @xmath21 .
note that in ( a ) @xmath83 and in ( b ) @xmath84 ; in both cases @xmath85 , @xmath86 .
the data points are the result of the numerical integration of le ( [ 2.1 ] ) for the relevant choices of @xmath8 and @xmath21 ; the dashed curves are the corresponding analytical expressions of eq .
( [ 2.4 ] ) .
[ f4.1],title="fig:",width=336 ] vs. @xmath11 in the asymmetric three - well potentials ( a ) @xmath81 and ( b ) @xmath82 for different @xmath21 .
note that in ( a ) @xmath83 and in ( b ) @xmath84 ; in both cases @xmath85 , @xmath86 .
the data points are the result of the numerical integration of le ( [ 2.1 ] ) for the relevant choices of @xmath8 and @xmath21 ; the dashed curves are the corresponding analytical expressions of eq .
( [ 2.4 ] ) .
[ f4.1],title="fig:",width=336 ] although the typical trajectories are by far the most probable being @xmath87 , still their contribution to the average transition time @xmath27 is negligible , as they reach the exit point in a quite short time , see eq .
( [ 2.7 ] ) .
by contrast , the barrier crossings may well be very unlikely
being @xmath77 exponentially small , but the particle , after falling into well @xmath31 , takes an exponentially long time of the order of @xmath48 [ see eq .
( [ 2.3 ] ) for @xmath88 , to recross into well @xmath13 .
the contribution to @xmath27 from such rare trajectories amounts to @xmath89 , that is , to our estimate in eq .
( [ 2.6 ] ) . in conclusion ,
as long as we characterize the relaxation in the overdamped potential @xmath8 by measuring the exit times , @xmath5 , the otherwise sporadic trajectories crossing the barrier may become _ dominant _ , depending on the injection point .
of course , this argument only applies for small , but finite noise strengths , i.e. , @xmath90 , whereas in the noiseless regime , @xmath91 , there exists only one allowed deterministic trajectory running downhill from @xmath11 to @xmath12 for any @xmath92 . in real or numerical experiments one can easily sample relaxation trajectories from @xmath11 to @xmath12 and distribute them according to their temporal length , @xmath5 .
based on the argument above , where the regular trajectories are regarded as much faster than the dominant ones , the @xmath72-distribution density , @xmath93 $ ] , can be separated into two distinct terms , i.e. , @xmath94 . for the sake of a comparison with actual data , in the regime of weak noise one can introduce the approximations , @xmath95,\ ] ] for the long exit times of the statistically rare trajectories crossing the barrier , and @xmath96^{\frac{1}{2 } } \cdot \exp[-({v_0-v_a})/{d(t)}],\ ] ] with @xmath97 , for the intrawell relaxation trajectories .
our expression for @xmath70 holds good for the harmonic approximation of the potential well @xmath13 , that is , by setting @xmath98 and ignoring all anharmonic terms of the third order and higher .
it was derived by standard mfpt methods @xcite and can be reformulated to match earlier solutions for @xmath72-distribution in a harmonic well @xcite . in eq .
( [ 3.4 ] ) for @xmath73 , we approximated the probability of barrier crossing as @xmath99 , and made use of the well - established exponential distribution for kramers escape times from @xmath31 back to @xmath13 @xcite . in fig .
[ f3.1 ] we display the outcome of an extensive numerical simulation of the exit process , eq .
( [ 2.1 ] ) , for the potential of fig .
[ f2.1 ] and different values of @xmath21 . as the injection point
is shifted past the threshold @xmath39 , also the relaxation time distributions change abruptly .
an exponential tail associated with the dominant trajectories becomes visible for @xmath100 ( inset ) ; as predicted in eq .
( [ 3.4 ] ) , such a tail has a small amplitude of the order of @xmath101 and decays slowly with time constant @xmath48 .
the distributions of the short relaxation times due to the regular trajectories , main panel , are reminiscent of the @xmath72-distributions in a harmonic well , @xmath70 of eq .
( [ ptharm ] ) .
however , the agreement gets quantitatively close only when @xmath11 approaches @xmath12 , the convergence being rather slow .
we attributed this inconvenience to the spatial asymmetry of well @xmath13 .
moreover , we remark that the average @xmath102 taken over the regular trajectories only , namely by using the approximate distribution density eq .
( [ ptharm ] ) , is a monotonic decreasing function of @xmath21 ; for vanishingly small @xmath21 values it comes close to the predicted estimate in eq .
( [ 2.7 ] ) .
the results of sec . [ bistable ] can be extended to study transitions in multiwell potentials , as well .
however , the algebraic manipulations on the mfpt ( [ 2.3 ] ) can become more complicated due to the multi - peaked structure of the functions @xmath43 and @xmath44 .
luckily , to gain a better understanding of the role of the dominant trajectories in the most general case of a disordered potential , it suffices to analyze in some detail the three - well potentials , only .
while any disordered potential can be regarded as an appropriate sequence of three - well potentials , it is clear that the relaxation properties discussed below only apply in the limit of infinite observation times , where the diffusing particle is allowed to explore the entire potential profile .
shorter observation times would necessarily restrict our analysis to the portion of the potential profile actually accessed by the particle .
let us imagine to add a third well to the potential plotted in fig .
[ f2.1 ] . if we agree on that the exit well must be at the bottom of the lowest one , then two geometries are possible , as illustrated in fig .
let @xmath103 and @xmath104 denote , respectively , the first and the second well to the right of well @xmath13 , with barriers @xmath105 and @xmath106 separating the three wells .
as for both wells @xmath107 , with @xmath86 , the equations @xmath108 may define two thresholds , @xmath109 , with @xmath110 . as a consequence ,
the barrier region of well @xmath13 is delimited from below by the threshold @xmath111 , that is , it starts at the level of the lower side - well @xmath112 see the geometric constructions in panels ( a ) and ( b ) .
now , the question rises naturally whether , in the presence of two ( or more ) thresholds , the slope of @xmath27 changes at each of them , and where such changes are possibly the most pronounced .
the answer is illustrated in the two panels of fig .
[ f4.1 ] , where the mfpt ( [ 2.3 ] ) has been plotted over an @xmath11 range comprising both @xmath109 : on reducing the noise intensity , a sharp crossover between a logarithmic and an exponential @xmath11 dependence emerges in the neighborhood of @xmath113 , whereas no substantial mfpt change can be associated with the other threshold .
this conclusion can be confirmed qualitatively by extending the semi - quantitative approach of sec .
[ bistable ] to both potentials of fig .
[ f4.1 ] . in the barrier region ,
the average transient time @xmath27 is dominated by the lower side - well @xmath112 ; the dominant trajectories cross one or two barriers , depending on which side - well is deeper .
accordingly , in the barrier region @xmath114 , the curve @xmath27 grows proportional to @xmath115/d)$ ] .
note that in view of the remark at the bottom of sec .
[ bistable ] , should one side - well @xmath116 sit lower than well @xmath13 , then such an exponential dependence would apply throughout the entire range @xmath117 and no logarithmic - to - exponential crossover would occur . vs. @xmath11 in the degenerate three - well potential @xmath118 for different @xmath21 ; simulation ( symbol ) versus analytical results ( dashed curves ) . note that @xmath119 and the threshold @xmath120 is defined by the condition @xmath121 , where @xmath122 ( see text ) .
the dashed curves were obtained by numerically integrating eq .
( [ 2.4 ] ) for the three - well potential .
[ f4.2 ] ] we consider now the special case of a three - well potential with two degenerate lower minima , say , in @xmath123 and @xmath124 , see fig .
this means that wells @xmath13 and @xmath103 are equally deep , while the third well sits higher up , that is , @xmath125 .
then , the process ( [ 2.1 ] ) models the relaxation occurring between two degenerate states , a mechanism often invoked in the chemical physical literature . as discussed in sec .
[ introduction ] , for low noise levels this problem is commonly addressed by ignoring the presence of more energetic states in the neighborhood .
however , the remarkable dependence of @xmath126 on the injection point , @xmath11 , shown in figure , suggests a different picture . as long as @xmath11 is confined around the bottom of well @xmath103 ,
the mfpt from @xmath11 to @xmath12 is almost independent of @xmath11 and well reproduced by the kramers rate of eq .
( [ 2.5 ] ) upon replacing @xmath11 with @xmath103 , and @xmath29 with @xmath105 .
in this case the role of well @xmath104 is irrelevant .
however , on moving @xmath11 to the right of a certain threshold @xmath39 , @xmath126 suddenly jumps up to a much higher value , insensitive to any further increase of @xmath11 .
the location of the threshold point @xmath127 and the magnitude of the mfpt jump can be explained as follows .
we assume that the lower @xmath126 plateau for @xmath128 is due to the regular trajectories crossing from @xmath103 to @xmath13 directly over barrier @xmath105 and , therefore , proportional to @xmath129 , whereas the higher plateau must come from those rare trajectories that cross first barrier @xmath106 to the right , with probability proportional to @xmath130 $ ] .
the time they take to cross back from well @xmath104 to well @xmath103 ( and then to well @xmath13 ) is a kramers s time proportional to @xmath131 .
therefore , their weighted contribution to the mfpt is proportional to @xmath132 and , most remarkably , supersedes the contribution from the regular trajectories for @xmath133 .
accordingly , @xmath39 is determined by choosing @xmath121 , where @xmath122 is the barrier height separating wells @xmath13 and @xmath103 see the geometric construction in fig .
[ f4.2 ] . as long as @xmath134
, the threshold @xmath39 is well defined .
therefore , there can exist a barrier region inside well @xmath103 , @xmath135 , such that the relaxation trajectories creeping into well @xmath104 are indeed _ dominant_. the corresponding @xmath72-distributions are well fitted by double exponential functions ( not shown ) with decay constants equal to the two plateau values of the curves @xmath126 versus @xmath11 .
many systems in condensed matter are described by an overdamped particle that diffuses on a disordered energy landscape of appropriate dimensionality , without ever reaching a proper equilibrium state ( glassy materials are a good example ) .
the physical chemical properties of these systems are often interpreted in terms of the relaxation rates inside single locally stable states or between pairs of locally stable states . however , determining such rates experimentally , through microscopic techniques , or even numerically , may prove a moot problem . as discussed in secs .
[ bistable ] and [ multiwell ] , the investigator who intends to proceed by weakly exciting the system out of its locally stable state and then letting it relax back to it , may encounter the difficulty of establishing whether the measured relaxation time depends on the presence of other metastable states .
this difficulty can be circumvented by a more restrictive definition of locally stable state .
our analysis clearly shows that in 1d the relaxation times within a single potential well or between degenerate wells can be determined by ignoring additional potential wells only under the condition that the energy of what we call the injection point is sufficiently close to the energy of the well bottom .
how close , it depends on the actual distribution of the wells along the potential landscape .
indeed , the critical threshold is determined by the lowest lying well , an information usually unavailable to the investigator .
therefore , above a certain ( but unknown ) threshold of the injection energy , the measured relaxation times exhibit a marked _ nonlocal _ dependence on the global potential profile .
such a nonlocal effect is due to the contribution from slower , though rare , relaxation trajectories , which explore the potential landscape surrounding the well(s ) of interest .
their presence can be appreciated , for instance , by looking at the distribution of the relevant relaxation times , though at the expense of much longer observation times .
the present analysis was restricted to 1d potentials for the sake of clarity , thus making our presentation hopefully easier to follow and affording higher numerical statistics .
its extension to potentials in two and even higher dimensions confirms the overall picture summarized here and is presently matter of further investigation .
we thank riken s ricc for computational resources .
y. li is supported by the nsf china under grant no .
p.k.g . is supported by serb start - up research grant ( young scientist ) no .
yss/2014/000853 and the ugc - bsr start - up grant no .
f.30 - 92/2015 m. goldstein , _ viscous liquids and the glass transition : a potential energy barrier picture _ , j. chem .
phys . * 51 * , 3728 ( 1969 ) .
h. frauenfelder , s. g. sligar , and p. g. wolynes , _ the energy landscapes and motions of proteins _ ,
science * 254 * 1598 ( 1991 ) .
d. l. malandro , and d. j. lacks , _ relationships of shear - induced changes in the potential energy landscape to the mechanical properties of ductile glasses _ , j. chem . phys . * 110 * , 4593 ( 1999 ) .
s. sastry , p. g. debenedetti , and f. h. stillinger , _ signatures of distinct dynamical regimes in the energy landscape of a glass - forming liquid _ , nature * 393 * , 554 ( 1998 ) .
g. adam and j. h. gibbs , _ on the temperature dependence of cooperative relaxation properties in glass - forming liquids _ , j. chem
. phys . * 43 * 139 ( 1965 ) .
p. hnggi , p. talkner , and m. borkovec , _ reaction - rate theory : fifty years after kramers _ , rev .
phys . * 62 * 251 ( 1990 ) .
l. luo and l .- h .
tang , _ sample - dependent first - passage - time distribution in a disordered medium _ , phys .
e * 92 * , 042137 ( 2015 ) .
g. r. bowman , v. s. pande , and f. no ( eds . ) , _ an introduction to markov state models and their application to long timescale molecular simulation _
( springer , dordrecht , 2014 ) .
m. ferrario , g. ciccotti , and k. binder ( eds . ) , _ physics computer simulations in condensed matter systems : from materials to chemical biology volume 1 _ , lect .
notes phys .
* 703 * ( springer , berlin heidelberg 2006 ) . c. r. doering , k. v. sargsyan , l. m. sander , and e.vanden-eijnden , _ asymptotics of rare events in birth - death processes bypassing the exact solutions _ , j. phys .
matter * 19 * , 065145 ( 2007 ) .
b. e. vugmeister , j. botina , and h. rabitz , _ nonstationary optimal paths and tails of prehistory probability density in multistable stochastic systems _ ,
e * 55 * , 5338 ( 1997 ) .
s. m. soskin , _ most probable transition path in an overdamped system for a finite transition time _ , p hys .
lett . a * 353 * , 281 ( 2006 ) , and additional references therein .
r. l. stratonovich , radiotekh .
( moscow ) * 3 * , 497 ( 1958 ) .
english translation in _ non - linear transformations of stochastic processes _ , edited by p. i. kuznetsov , r. l. stratonovich , and v. i. tikhonov ( pergamon , oxford , 1965 ) . c. w. gardiner _
handbook of stochastic methods _
( springer , berlin , 1985 ) . chapters 5 and 9 .
n. goel and n. richter - dyn , _ stochastic models in biology _
( academic press , new york , 1974 ) .
a. szabo , k. schulten , and z. schulten , _ first passage time approach to diffusion controlled reactions _
, j. chem .
phys . * 72 * , 4350 ( 1980 ) .
z. hu , l. cheng , and b. j. berne , _ first passage time distribution in stochastic processes with moving and static absorbing boundaries with application to biological rupture experiments _
, j. chem .
phys . * 133 * , 034105 ( 2010 ) . | we investigate both analytically and by numerical simulation the relaxation of an overdamped brownian particle in a 1d multiwell potential .
we show that the mean relaxation time from an injection point inside the well down to its bottom is dominated by statistically rare trajectories that sample the potential profile outside the well . as a consequence ,
also the hopping time between two degenerate wells can depend on the detailed multiwell structure of the entire potential .
the nonlocal nature of the transitions between two states of a disordered landscape is important for the correct interpretation of the relaxation rates in complex chemical - physical systems , measured either through numerical simulations or experimental techniques . |
spontaneous breaking of lepton number leads to a massless goldstone boson , the majoron ( @xmath0 ) @xcite .
there are two well - known experimental probes for the majoron : the first is the invisible width of the @xmath4 boson , very precisely measured at lep @xcite .
the second is neutrinoless double beta decay @xcite .
the nemo-3 collaboration , for example , has published limits on half - lives for majoron - emitting neutrinoless double beta decay for a number of isotopes @xcite .
in addition , there are different astrophysical constraints on the majoron from the cooling of red giant stars and supernovae @xcite .
another interesting possibility to search for majorons , namely charged lepton decays with majoron emission , has attracted considerably less attention .
indeed , the limits on @xmath5 quoted by the particle data group @xcite are all based on experimental data which is now more than 20 years old .
probably this apparent lack of interest from the experimental side is due to the fact that both , the triplet @xcite and the doublet majoron @xcite , are ruled out by lep data , while the ( classical ) singlet majoron model @xcite predicts majoron - neutrino and majoron - charged - lepton couplings which are unmeasurably small .
supersymmetry ( susy ) is a theoretically well motivated extension of the standard model . with the first data taking of the lhc only months away , searches for susy will gain momentum soon .
one of the many virtues of susy is the fact that the minimal supersymmetric extension of the standard model ( mssm ) provides an interesting candidate for the cold dark matter ( cdm ) , usually assumed to be the lightest neutralino ( @xmath6 ) _ if r - parity is conserved . _ a stable , electrically neutral lightest supersymmetric particle ( lsp ) will escape the detector and lead to the famous missing momentum signal , upon which standard susy searches are based .
if r - parity is broken , the lsp decays and the cdm candidate is lost . for explicit r - parity violation ( ) , the missing energy signal is degraded , but a larger number of jets and charged lepton final states should make discovery of a ( comparatively ) easy task . in _ spontaneous _ ( s- ) , however , the lightest neutralino can decay invisibly through @xmath7 .
as pointed out in @xcite , if the scale of is low , this decay mode can be easily dominant and s - can be confused with a standard mssm with @xmath8 conserved . here
, we revisit @xmath5 within spontaneous r - parity violation .
our calculation is based on the model of @xcite . in this model
the majoron is mainly singlet , thus escaping the lep bounds .
this is different from the original spontaneous model @xcite , which used the left - sneutrinos to break r - parity .
nevertheless , in our model the majoron can play an important role phenomenologically . in @xcite @xmath5
was calculated for a tau neutrino mass of @xmath9 mev . here
we show that ( a ) despite the fact that current neutrino mass bounds are of the order of ev or less , theoretically @xmath10 can be ( nearly ) arbitrarily large in s- , and ( b ) @xmath10 is large in the same part of susy parameter space where the invisible neutralino decay is large , making the discovery of r - parity violation at the lhc difficult .
br(@xmath10 ) thus gives complementary information to accelerator experiments . at the same time , the meg experiment @xcite has started taking data .
meg is optimised to search for br(@xmath11 ) with a sensitivity of br(@xmath11 ) @xmath12 ( few ) @xmath13 .
while the impressive statistics of the experiment should allow , in principle , to improve the existing bound on br(@xmath10 ) @xcite by a considerable margin , the experimental triggers and cuts make it necessary to resort to a search for the radiative majoron emission mode , br(@xmath14 ) , if one wants to limit ( or measure ) the majoron - charged - lepton coupling .
we therefore also calculate br(@xmath14 ) . in the next section
, we will briefly discuss s - and give an approximative , analytical estimation of br(@xmath5 ) and br(@xmath15 ) .
we then present our numerical results in section ( [ sec : num ] ) , showing the correlation of br(@xmath5 ) with the invisible branching ratio of the lightest neutralino decay . in section ( [ sec : exp ] ) we then turn to a brief discussion of existing experimental bounds and comment on how a limit on br(@xmath14 ) might be used to put a bound on the majoron - charged - lepton coupling .
we then close with a brief conclusion in section ( [ sec : cncl ] ) .
the model we consider @xcite extends the particle spectrum of the mssm by three additional singlet superfields , @xmath16 , @xmath17 and @xmath18 , with lepton number assignments of @xmath19 respectively .
the superpotential can be written as @xmath20 strictly speaking only @xmath16 is necessary to spontaneously break @xmath8 .
the inclusion of @xmath17 and @xmath18 allows to construct a superpotential which purely consists of trilinear terms , thus potentially solving also the @xmath21 problem of the mssm . at low energy various fields
acquire vacuum expectation values ( vevs ) .
besides the usual mssm higgs boson vevs @xmath22 and @xmath23 , these are @xmath24 , @xmath25 , @xmath26 and @xmath27 .
note , that @xmath28 generates effective bilinear terms @xmath29 and that @xmath30 , @xmath31 and @xmath32 violate lepton number as well as r - parity .
the observed smallness of neutrino masses guarantees that the operators generated by these vevs are small .
the smallness of @xmath33 implies that either @xmath34 or @xmath30 is small , but not necessarily both . details of the model , such as mass matrices and couplings , can be found in @xcite , for the phenomenology of the lsp decay in this model see @xcite . for brevity in the following
we will concentrate on only a few , relevant aspects of the phenomenology of this model : the neutrino mass matrix , the lightest neutralino decay to majorons and charged lepton decays .
the main motivation to study r - parity breaking supersymmetry certainly is that generates neutrino masses and thus contains a possible explanation for the observed neutrino oscillation data . for the spontaneous model defined in eq .
( [ eq : wsuppot ] ) , the effective neutrino mass matrix at tree - level can be cast into a very simple form @xmath35 here , @xmath36 , with @xmath37 .
the coefficients @xmath38 , @xmath39 and @xmath40 are defined as @xmath41 @xmath42 is the determinant of the ( @xmath43 ) matrix of the heavy neutral states ( the four mssm states , @xmath44 , @xmath45 and @xmath46 , plus the three fermionic components of the new singlet superfields of eq .
( [ eq : wsuppot ] ) ) @xmath47\ ] ] and @xmath48 .
the `` photino '' mass parameter is defined as @xmath49 . since @xmath37 @xmath50 in the limit of large @xmath51 .
one can easily fit the observed neutrino masses and angles using eq .
( [ eq : eff ] ) , see @xcite and the short discussion in the next section . from a phenomenological point of view
the most important difference between spontaneous and explicit r - parity violating models is the appearance of the majoron .
the pseudo - scalar sector of the model we consider has eight different eigenstates .
two of them are goldstone bosons .
the standard one is eaten by the @xmath4 boson , the remaining state is identified with the majoron . in the limit @xmath52 the majoron profile is given by the simple expression @xmath53 here , @xmath54 and terms of order @xmath55 , where @xmath56 , have been neglected .
majorons are weakly coupled , thus potentially lead to a decay mode for the lightest neutralino which is invisible .
neutralino - majoron couplings can be calculated from the general coupling @xmath57 @xmath58 mixing between the neutralinos and the neutrinos then leads to a coupling @xmath59 . in the limit
@xmath60 one can derive a very simple approximation formula for @xmath61 .
it s given by @xcite @xmath62 where the dots stand for higher order terms neglected here and @xmath63 is the matrix which diagonalizes the ( mssm ) neutralino mass matrix . @xmath64 and @xmath65 . here
@xmath66 is the matrix which diagonalizes either the part of the ( @xmath67 ) effective neutrino mass matrix , proportional to @xmath38 or @xmath40 , depending on which gives the larger eigenvalue .
( [ eq : majcl ] ) shows that for constant @xmath68 and @xmath69 , @xmath70 as @xmath30 goes to infinity .
this is as expected , since for @xmath71 the spontaneous model approaches the explicit bilinear model .
we note that , in addition to the majoron there is also a rather light singlet scalar , called the `` scalar partner '' of the majoron in @xcite , @xmath72 .
the lightest neutralino has a coupling @xmath73 , which is of the same order as @xmath61 .
since @xmath72 decays to nearly 100 @xmath74 to two majorons , this decay mode contributes sizeably to the invisble width of the lightest neutralino , for more details see @xcite .
the decays @xmath5 can be calculated from the general coupling @xmath75 . in the limit of small r - parity violating parameters the relevant interaction lagrangian for the @xmath76 coupling
is given by @xmath77 with @xmath78\nonumber \\
o_{lijj}^{ccp } & = & \big(o_{rjij}^{ccp}\big)^*.\end{aligned}\ ] ] the @xmath79 coefficients are different combinations of mssm parameters @xmath80 where @xmath81 is the determinant of the mssm chargino mass matrix @xmath82 .
( [ cpl_llj ] ) shows that one expects large partial widths to majorons , if @xmath30 is low . for a charged lepton @xmath83 , with polarization vector @xmath84 ,
the decay @xmath85 has a differential decay width given by @xmath86 \nonumber \label{eq : dgampol}\end{aligned}\ ] ] where @xmath87 is the angle between the polarization vector @xmath84 and the momentum @xmath88 of the charged lepton in the final state , and @xmath89 is the polarization degree of the decaying charged lepton . in the limit @xmath90 the expression simplifies to @xmath91 since @xmath92 .
the angular distribution of the majoron emitting lepton decay is thus very similar to the standard model muon decay @xcite , up to corrections of the order @xmath93 , which are negligible in practice .
we next consider the decay @xmath2 .
can be found from straightforward replacements . ]
it is induced by the feyman diagrams shown in fig .
( [ fig : feyng ] ) . .
as in the standard model radiative decay @xmath94 these diagrams contain an infrared divergence for @xmath95 , see text.,title="fig:",width=188,height=113 ] . as in the standard model radiative decay @xmath94
these diagrams contain an infrared divergence for @xmath95 , see text.,title="fig:",width=188,height=113 ] in the approximation @xmath96 the partial decay width for the process @xmath2 can be written as @xmath97 where @xmath98 is a phase space integral given by @xmath99 the dimensionless parameters @xmath100 , @xmath101 are defined as usual @xmath102 and @xmath103 and @xmath104 are the minimal electron and photon energies measured in a given experiment .
note that the integral @xmath98 diverges for @xmath105 .
this infrared divergence is well - known from the standard model radiative decay @xmath94 , and can be taken care off in the standard way by introducing a non - zero photon mass @xmath106 .
note that in the limit @xmath107 there also appears a colinear divergence , just as in the sm radiative decay . since in any practical experiment there is a minimum measurable photon energy , @xmath104 , as well as a minimum measurable photon - electron angle ( @xmath108 ) , neither divergence affects us in practice .
we simply integrate from the minimum value of @xmath101 up to @xmath109 when estimating the experimental sensitiviy of br(@xmath2 ) on the majoron coupling . in the calculation of the integral @xmath98 one
has to take into account not only the experimental cuts applied to the variables @xmath100 and @xmath101 , but also the experimental cut for the angle between the directions of electron and photon .
this angle is fixed for kinematical reasons to @xmath110 this relation restricts @xmath111 to be @xmath112 as a function of @xmath101 ( and vice versa ) and to @xmath113 for @xmath114 . using the formula for @xmath115 , in the approximation @xmath96
, @xmath116 one finds a very simple relation between the two branching ratios @xmath117 we will use eq . ( [ eq : relbr ] ) in section ( [ sec : exp ] ) when we discuss the relative merits of the two different measurements .
all numerical results shown in this section have been obtained using the program package spheno @xcite , extended to include the new singlet superfields @xmath16 , @xmath17 and @xmath18
. we always choose the parameters in such a way that solar and atmospheric neutrino data @xcite are fitted correctly .
the numerical procedure to fit neutrino masses is the following .
for any random choice of mssm parameters , we can reproduce the `` correct '' mssm value of @xmath21 for a random value of @xmath51 , by appropriate choice of @xmath118 . for any random set of @xmath119 , @xmath120 , @xmath31 and @xmath30
, we can then calculate those values of @xmath34 and @xmath32 , using eq .
( [ eq : eff ] ) , such that the corresponding @xmath33 and @xmath121 give correct neutrino masses and mixing angles . in the plots shown below we use @xmath121 for the atmospheric scale and @xmath33 for the solar scale .
as shown previously @xcite if the lightest neutralino is mainly a bino , the decay to majoron plus neutrino is dominant if @xmath30 is low .
this is demonstrated again for a bino lsp in fig .
( [ fig : vis ] ) , to the left , for a sample point using msugra parameters @xmath122 gev , @xmath123 gev , @xmath124 , @xmath125 gev and sgn@xmath126 .
we stress that this result is independent of the choice of msugra parameters to a large degree @xcite .
a scan over @xmath51 has been performed in this plot , varying @xmath51 in the huge interval [ @xmath127 tev .
large values of @xmath51 lead to small values of the constant @xmath40 in the neutrino mass matrix , see eq .
( [ eq : eff ] ) .
small @xmath40 require , for constant neutrino masses , large values of @xmath33 , which in turn lead to a large invisible width of the neutralino .
the largest values of @xmath51 ( dark areas ) therefore lead to the smallest visible neutralino decay branching ratios shown in fig .
( [ fig : vis ] ) . ) ( right ) versus @xmath30 in gev for a number of different choice of @xmath51 between [ @xmath127 tev indicated by the different colours .
darker colours indicate larger @xmath51 in a logarithmic scale .
msugra parameters defined in the text .
there is very little dependence on the actual msugra parameters , however , see discussion and fig .
( [ fig : visvmuej]).,title="fig:",width=302,height=226 ] ) ( right ) versus @xmath30 in gev for a number of different choice of @xmath51 between [ @xmath127 tev indicated by the different colours .
darker colours indicate larger @xmath51 in a logarithmic scale .
msugra parameters defined in the text .
there is very little dependence on the actual msugra parameters , however , see discussion and fig .
( [ fig : visvmuej]).,title="fig:",width=302,height=226 ] fig .
( [ fig : vis ] ) , to the right , shows the branching ratio br(@xmath10 ) as a function of @xmath30 for different values of @xmath51 .
all parameters have been fixed to the same values as shown in the left figure .
as the figure demonstrates , small values of @xmath30 ( and large values of @xmath51 ) lead to large values of br(@xmath10 ) .
this agrees with the analytic expectation , compare to eq .
( [ cpl_llj ] ) .
our main result is shown in fig .
( [ fig : visvmuej ] ) . in this figure
we show br(@xmath3 ) versus the sum of all branching ratios of neutralino decays leading to at least one visible particle in the final state for two different choices of msugra parameters .
the similarity of the two plots shows that our result is only weakly dependent on the true values of msugra parameters .
we have checked this fact also by repeating the calculation for other msugra points , although we do not show plots here . as expected br(@xmath10 ) anticorrelates with the visible bino decay branching fraction and thus probes a complementary part in the supersymmetric parameter space .
an upper bound on br(@xmath10 ) will constrain the maximum branching ratio for invisible neutralino decay , thus probing the part of parameter space where spontaneous r - parity breaking is most easily confused with _ conserved _
r - parity at accelerators . ) for two msugra points , for various choices of @xmath51 , see fig .
( [ fig : vis ] ) . to the left ,
same msugra parameters as fig .
( [ fig : vis ] ) , to the right sps1a.,title="fig:",width=302,height=226 ] ) for two msugra points , for various choices of @xmath51 , see fig .
( [ fig : vis ] ) . to the left ,
same msugra parameters as fig .
( [ fig : vis ] ) , to the right sps1a.,title="fig:",width=302,height=226 ] we have checked the points shown in the plots for various phenomenological constraints .
lep bounds are trivially fulfilled by @xmath128 .
double beta decay bounds on @xmath129 when discussing experimental bounds , to differentiate from the model dependent couplings @xmath130 and @xmath131 defined in section ( [ sec : th ] ) . ]
are of the order of @xmath132 @xcite and , since the coupling @xmath129 is suppressed by two powers of r - parity violating parameters , are easily satisfied in our model .
more interesting is the astrophysical limit on @xmath133 .
@xcite quotes a bound of @xmath134 .
although this bound is derived from the coupling of the majoron to two electrons , thus constraining actually the products @xmath135 , @xmath136 and @xmath137 , whereas br(@xmath10 ) is proportional to @xmath138 and @xmath139 , it still leads to a ( weak ) constraint on br(@xmath3 ) , since neutrino physics shows that two leptonic mixing angles are large .
this requires that either @xmath140 _ or _ @xmath141 . for the case studied in our plots , where @xmath33 generate the solar scale ,
@xmath142 requires @xmath143 .
numerically we then find that @xmath134 corresponds to an upper bound on br(@xmath1 ) of very roughly br(@xmath1 ) @xmath144 . in case
the neutrino data is fitted with @xmath145 for the atmospheric scale , the corresponding bound is considerably weaker .
the particle data group @xcite cites @xcite with an upper limit on the branching ratio of br(@xmath147 ) @xmath148 , where @xmath149 is a scalar boson called the familon .
this constraint does not apply to the majoron we consider here , since it is derived from the decay of polarized muons in a direction opposite to the direction of polarization .
the authors of @xcite concentrated on this region , since it minimizes events from standard model @xmath150-decay . as shown in eq .
( [ eq : dgampol ] ) , the majoron emitting decay has a very similar angular distribution as the standard model decay , with the signal approaching zero in the data sample analyzed by @xcite .
nevertheless , from the spin processed data shown in fig.(7 ) of @xcite , which seems to be in good agreement with the sm prediction , it should in principle be possible to extract a limit on br(@xmath10 ) . from this figure we estimate very roughly that this limit should be about one order of magnitude less stringent than the one for familon decay . for a better estimate a re - analysis of this data ,
including systematic errors , would be necessary .
@xcite searched for majorons in the decay of @xmath151 , deriving a limit of br(@xmath151 ) @xmath152 . since the experimental cuts used in this paper @xcite are designed to reduce the standard model background from the decay chain @xmath153 , the contribution from on - shell muons is reduced by about five orders of magnitude .
the limit then essentially is a limit on the majoron - neutrino - neutrino coupling , @xmath129 , leaving only a very weak constraint on the coupling @xmath154 .
also an analysis searching for br(@xmath146 ) has been published previously @xcite . from a total data sample of @xmath155 stopped muons over the live time of the experiment @xcite derived a limit on br(@xmath146 ) of the order of br(@xmath146 ) @xmath156 . for the cuts used in this analysis , we calculate @xmath157 .
thus , see eq.([eq : relbr ] ) , this limit translates into only a rather weak bound br(@xmath3 ) @xmath158 .
the meg experiment @xcite has a muon stopping rate of @xmath159 per second and expects a total of the order of @xmath160 muons over the expected live time of the experiment .
an analysis of _ electron only _ events near the endpoint should therefore allow , in principle , to improve the existing limits on br(@xmath3 ) by an estimated ( @xmath161 ) orders of magnitude , if systematic errors can be kept under control .
however , the meg experiment , as it is designed to search for br(@xmath162 ) , uses a trigger that requires a photon in the event with a minimum energy of @xmath163 mev .
meg data can thus constrain the majoron - charged - lepton coupling only via searching for @xmath164(@xmath146 ) . as a function of @xmath103 for three different values of @xmath165 from top to bottom and for two different values of @xmath166 . to the left @xmath167 , to the right @xmath168.,title="fig:",width=302,height=226 ] as a function of @xmath103 for three different values of @xmath165 from top to bottom and for two different values of @xmath166 . to the left @xmath167 , to the right @xmath168.,title="fig:",width=302,height=226 ] fig .
( [ fig : int ] ) shows the value of the phase space integral @xmath98 as a function of @xmath103 for three different values of @xmath104 and for two choices of @xmath166 .
the meg proposal describes the cuts used in the search for @xmath11 as @xmath169 , @xmath170 and @xmath171 8.4 mrad . for these values
we find a value of @xmath172 .
a limit for br(@xmath11 ) of br(@xmath11 ) @xmath173 then translates into a limit of br@xmath174 , obviously not competitive . to improve upon this bound , it is necessary to relax the cuts .
for example , relaxing the cut on the opening angle to @xmath167 , the value of the integral increases by more than 3 orders of magnitude for @xmath175 .
on the other hand , such a change in the analysis is prone to induce background events , which the meg cuts were designed for to avoid .
the meg proposal discusses as the two most important sources of background : ( a ) prompt events from the standard model radiative decay @xmath176 ; and ( b ) accidental background from muon annihilation in flight . for the current experimental setup
the accidental background is larger than the prompt background .
certainly , a better timing resolution of the experiment would be required to reduce this background . for the prompt background we estimate , using the formulas of @xcite , that for a total of @xmath177 muon events , one background event from the radiative decay will enter the analysis window for @xmath178 for the current cut on @xmath166 .
a further relaxation of the cuts can lead , in principle , to much larger values for @xmath98 .
however , the search for br@xmath179 than necessarily is no longer background free . since all the events from @xmath180 lie along the line of @xmath166 defined by eq .
( [ eq : ctheta ] ) , whereas events from the sm radiative mode fill all of the @xmath166 space , such a strategy might be advantageous , given a large enough data sample . before closing this section , we mention that tau decays with majoron emssion are less interesting phenomenologically for two reasons .
first the existing experimental limits are much weaker for taus @xcite br(@xmath181 ) @xmath182 @xmath74 and br(@xmath183 ) @xmath184 @xmath74 . and , second
, although the coupling @xmath185 is larger than the coupling @xmath186 by a factor @xmath187 , the total width of the tau is much larger than the width of the muon , thus the resulting theoretical predictions for tau branching ratios to majorons are actually smaller than for the muon by a factor of approximately @xmath188 .
we have calculated branching ratios for exotic muon and tau decays involving majorons in the final state .
branching ratios can be measurably large , if the scale of lepton number breaking is low .
this result is independent of the absolute value of the neutrino mass .
the lowest possible values of @xmath30 ( at large values of @xmath51 ) are already explored by the existing limit on br(@xmath10 ) .
we have briefly discussed the status of experimental limits .
it will not be an easy task to improve the current numbers in future experiments . while meg @xcite certainly has a high number of muon events in the detector , a search for br@xmath179 instead of br@xmath189 suffers from a small value of the available phase space integral , given current meg cuts .
an improvement will only be possible , if a dedicated search by the experimentalists is carried out .
nevertheless , we believe this is a worthwhile undertaking , since measuring a finite value for br(@xmath10 ) will establish that r - parity is broken in a region of susy parameter space complementary to that probed by accelerator searches .
work supported by spanish grants fpa2008 - 00319/fpa and accion integrada no ha-2007 - 0090 ( mec ) .
thanks the generalitat valenciana for financial support . w.p . is supported by the daad , project number d/07/13468 , and partially by the german ministry of education and research ( bmbf ) under contract 05ht6wwa .
y. chikashige , r. n. mohapatra and r. d. peccei , phys .
b * 98 * , 265 ( 1981 ) .
g. b. gelmini and m. roncadelli , phys .
b * 99 * , 411 ( 1981 ) . c. s. aulakh and r. n. mohapatra , phys .
b * 119 * , 136 ( 1982 ) . c. amsler _ et al .
_ [ particle data group ] , phys .
b * 667 * , 1 ( 2008 ) .
majoron emission in neutrinoless double beta decay has been first discussed in : h. m. georgi , s. l. glashow and s. nussinov , nucl .
b * 193 * , 297 ( 1981 ) ; for reviews on double beta decay , see for example : m. doi , t. kotani and e. takasugi , prog .
suppl . * 83 * , 1 ( 1985 ) ; w. c. haxton and g. j. stephenson , prog .
* 12 * ( 1984 ) 409 .
r. arnold _ et al . _
[ nemo collaboration ] , nucl .
a * 765 * , 483 ( 2006 ) [ arxiv : hep - ex/0601021 ] ; for updated information on the experiment see , for example : l. simard , arxiv:0810.0533 [ hep - ex ] . g. g. raffelt , phys . rept .
* 198 * , 1 ( 1990 ) ; g. g. raffelt , `` stars as laboratories for fundamental physics : the astrophysics of neutrinos , axions , and other weakly interacting particles , '' _ chicago , usa : univ
( 1996 ) 664 p _ ; g. g. raffelt , ann .
nucl . part .
sci . * 49 * , 163 ( 1999 ) [ arxiv : hep - ph/9903472 ] .
m. kachelriess , r. tomas and j. w. f. valle , phys .
d * 62 * , 023004 ( 2000 ) [ arxiv : hep - ph/0001039 ] .
m. hirsch , a. vicente and w. porod , phys .
d * 77 * , 075005 ( 2008 ) [ arxiv:0802.2896 [ hep - ph ] ] .
a. masiero and j. w. f. valle , phys .
b * 251 * , 273 ( 1990 ) .
j. c. romao , n. rius and j. w. f. valle , nucl .
b * 363 * , 369 ( 1991 ) .
proposal to psi : `` meg : search for @xmath190 down to @xmath13 branching ratio '' ; documents and status at http://meg.web.psi.ch/. for a status report see , for example : a. maki , aip conf .
* 981 * ( 2008 ) 363 .
m. hirsch , j. c. romao , j. w. f. valle and a. villanova del moral , phys . rev . *
d70 * , 073012 ( 2004 ) m. hirsch , j. c. romao , j. w. f. valle and a. villanova del moral , phys . rev . *
d73 * , 055007 ( 2006 ) w. porod , comput .
commun .
* 153 * , 275 ( 2003 ) m. maltoni , t. schwetz , m. a. tortola and j. w. f. valle , new j. phys . *
6 * , 122 ( 2004 ) ; an updated analysis with data included up to july 2008 has been published in t. schwetz , m. tortola and j. w. f. valle , arxiv:0808.2016 [ hep - ph ] .
a. jodidio _ et al .
_ , phys .
d * 34 * , 1967 ( 1986 ) [ erratum - ibid .
d * 37 * , 237 ( 1988 ) ] . c. e. picciotto _ et al .
_ , phys . rev .
d * 37 * , 1131 ( 1988 ) .
j. t. goldman _ et al .
_ , phys .
d * 36 * , 1543 ( 1987 ) .
y. kuno and y. okada , rev .
phys . * 73 * , 151 ( 2001 ) [ arxiv : hep - ph/9909265 ] .
r. m. baltrusaitis _ et al . _ [ mark - iii collaboration ] , phys .
* 55 * , 1842 ( 1985 ) . | in models where the breaking of lepton number is spontaneous a massless goldstone boson , the majoron ( @xmath0 ) , appears . we calculate the theoretically allowed range for the branching ratios of majoron emitting charged lepton decays , such as br(@xmath1 ) and br(@xmath2 ) , in a supersymmetric model with spontaneous breaking of r - parity .
br(@xmath3 ) is maximal in the same region of parameter space for which the lightest neutralino decays mainly invisibly .
a measurement of br(@xmath3 ) thus potentially provides information on r - parity violation complementary to accelerator searches .
we also briefly discuss existing bounds and prospects for future improvements on the majoron coupling to charged leptons . |
approaches to the late - time acceleration of the universe may be divided into three broad classes .
first , it is possible that there is some as yet undiscovered property of our existing model of gravity and matter that leads to acceleration at the current epoch . into this category
one might include the existence of a tiny cosmological constant and the possibility that the backreaction of cosmological perturbations might cause self - acceleration .
second is the idea that there exists a new dynamical component to the cosmic energy budget .
this possibility , with the new source of energy density modeled by a scalar field , is usually referred to as _ dark energy_. finally , it may be that curvatures and length scales in the observable universe are only now reaching values at which an infrared modification of gravity can make itself apparent by driving self - acceleration @xcite .
it is this possibility that i will briefly describe in this article , submitted to the proceedings of the nasa workshop _ from quantum to cosmos : fundamental physics research in space_. while i will mention a number of different approaches to modified gravity , i will concentrate on laying out the central challenges to constructing a successful modified gravity model and on illustrating them with a particular simple example . detailed descriptions of some of the other possible ways to approach this problem can be found in the excellent contributions of sean carroll , cedric deffayet , gia dvali and john moffat .
although , within the context of general relativity ( gr ) , one does nt think about it too often , the metric tensor contains , in principle , more degrees of freedom than the usual spin-2 _ graviton _ ( see sean carroll s talk in these proceedings for a detailed discussion of this ) .
the reason why one does nt hear of these degrees of freedom in gr is that the einstein - hilbert action is a very special choice , resulting in second - order equations of motion , which constrain away the scalars and the vectors , so that they are non - propagating .
however , this is not the case if one departs from the einstein - hilbert form for the action .
when using any modified action ( and the usual variational principle ) one inevitably frees up some of the additional degrees of freedom .
in fact , this can be a good thing , in that the dynamics of these new degrees of freedom may be precisely what one needs to drive the accelerated expansion of the universe .
however , there is often a price to pay .
the problems may be of several different kinds .
first , there is the possibility that along with the desired deviations from gr on cosmological scales , one may also find similar deviations on solar system scales , at which gr is rather well - tested .
second is the possibility that the newly - activated degrees of freedom may be badly behaved in one way or another ; either having the wrong sign kinetic terms ( ghosts ) , and hence being unstable , or leading to superluminal propagation , which may lead to other problems .
these constraints are surprisingly restrictive when one tries to create viable modified gravity models yielding cosmic acceleration . in the next few sections
i will describe several ways in which one might modify the action , and in each case provide an explicit , clean , and simple example of how cosmic acceleration emerges .
however , i will also point out how the constraints i have mentioned rule out these simple examples , and mention how one must complicate the models to recover viable models .
the simplest way one could think to modify gr is to replace the einstein - hilbert lagrangian density by a general function @xmath0 of the ricci scalar @xmath1 @xcite .
s = d^4 x + d^4 x l_m[_i , g _ ] , [ jordanaction ] where @xmath2 is the ( reduced ) planck mass and @xmath3 is the lagrangian density for the matter fields @xmath4 . here
, i have written the matter lagrangian as @xmath5 $ ] to make explicit that in this frame - the _ jordan _ frame - matter falls along geodesics of the metric @xmath6 .
the equation of motion obtained by varying the action ( [ jordanaction ] ) is ( 1+f_r ) r _ - g_(r+f ) + ( g_- _ _ ) f_r = , [ jordaneom ] where i have defined @xmath7 .
further , if the matter content is described as a perfect fluid , with energy - momentum tensor , @xmath8 where @xmath9 is the fluid rest - frame four - velocity , @xmath10 is the energy density and @xmath11 is the pressure , then the fluid equation of motion is the usual continuity equation .
when considering the background cosmological evolution of such models , i will take the metric to be of the flat robertson - walker form , @xmath12 . in this case , the usual friedmann equation of gr is modified to become 3h^2 -3f_r ( h+h^2)+f+18f_rrh(h+4hh)= [ jordanfriedmann ] and the continuity equation is _
m + 3h(_m+p_m)=0 .
[ jordancontinuity ] when supplied with an equation of state parameter @xmath13 , the above equations are sufficient to solve for the background cosmological behavior of the space - time and it s matter contents . for appropriate choices of the function
@xmath0 it is possible to obtain late - time cosmic acceleration without the need for dark energy , although evading bounds from precision solar - system tests of gravity turns out to be a much trickier matter , as we shall see . while one can go ahead and analyze this theory in the jordan frame , it is more convenient to perform a carefully - chosen conformal transformation on the metric , in order to render the gravitational action in the usual einstein hilbert form of gr . following the description in @xcite , consider the conformal transformation _ =
( x^ ) g _ ,
[ conftrans ] and construct the function @xmath14 that satisfies 1+f_r[r()]= . defining a rescaled scalar field by @xmath15 , with @xmath16 , the resulting action becomes = d^4 x r & + & d^4 x + & + & d^4 x e^-2 l_m[_i , e^-g _ ] , [ einsteinaction ] where the potential @xmath17 is determined entirely by the original form ( [ jordanaction ] ) of the action and is given by v()=\{e^r [ ( ) ] - f(r [ ( ) ] ) } .
[ einsteinpotential ] the equations of motion in the einstein frame are much more familiar than those in the jordan frame , although there are some crucial subtleties . in particular , note that in general , test particles of the matter content @xmath4 do not freely fall along geodesics of the metric @xmath18 . the equations of motion in this frame are those obtained by varying the action with respect to the metric @xmath18 _ = ( t _ + t^ ( ) _ ) , [ einsteineom ] with respect to the scalar field @xmath19 = - ( ) , [ scalareom ] and with respect to the matter fields @xmath4 , described as a perfect fluid .
once again , i will specialize to consider background cosmological evolution in this frame .
the einstein - frame line element can be written in familiar frw form as ds^2 = -dt^2+a^2(t)d*x*^2 , [ einsteinfrwmetric ] where @xmath20 and @xmath21 .
the einstein - frame matter energy - momentum tensor is then given by _
^m = ( _ m + p_m)u _ u _ + p_m g _ , [ einsteinperfectfluid ] where @xmath22 , @xmath23 and @xmath24
. for definiteness and simplicity focus on the simplest correction to the einstein - hilbert action ; @xmath25 , with @xmath26 a new parameter with units of @xmath27 $ ] . the field equation for the metric
is then [ cdtteqn ] ( 1+)r _ - ( 1-)rg _ + ^4r^-2 = . the constant - curvature vacuum solutions , for which @xmath28 , satisfy @xmath29 .
thus , there exists a constant - curvature vacuum solution which is de sitter space .
we will see that the de sitter solution is , in fact , unstable , albeit with a very long decay time @xmath30 .
the time - time component of the field equations for this metric is [ newfriedmann ] 3h^2 - ( 2hh + 15h^2h+2h^2 + 6h^4 ) = .
as i have discussed , one may now transform to the _ einstein frame _ , where the gravitational lagrangian takes the einstein - hilbert form and the additional degree of freedom appears as a fictitious scalar field @xmath19 , with potential @xmath31 shown in the figure below .
[ potential ] , title="fig : " ] denoting with a tilde all quantities ( except @xmath19 ) in the einstein frame , the relevant einstein - frame cosmological equations of motion are @xmath32 \ , \\\ ] ] @xmath33 where a prime denotes @xmath34 , and where @xmath35 \ , \ ] ] with @xmath36 a constant , and @xmath37 finally , note that the matter - frame hubble parameter @xmath38 is related to that in the einstein frame @xmath39 by @xmath40 how about cosmological solutions in the einstein frame ? ordinarily , einstein gravity with a scalar field with a minimum at @xmath41 would yield a minkowski vacuum state . however , here this is no longer true .
even though @xmath42 as @xmath43 , this corresponds to a curvature singularity and so is not a minkowski vacuum .
the other minimum of the potential , at @xmath44 , does not represent a solution .
focusing on vacuum solutions , i.e. , @xmath45 , the beginning of the universe corresponds to @xmath46 and @xmath47 .
the initial conditions we must specify are the initial values of @xmath19 and @xmath48 , denoted as @xmath49 and @xmath50 .
there are then three qualitatively distinct outcomes , depending on the value of @xmath50 .
_ 1 . eternal de sitter .
_ there is a critical value of @xmath51 for which @xmath19 just reaches the maximum of the potential @xmath17 and comes to rest . in this case
the universe asymptotically evolves to a de sitter solution ( ignoring spatial perturbations ) . as we have discovered before ( and is obvious in the einstein frame ) , this solution requires tuning and is unstable .
power - law acceleration .
_ for @xmath52 , the field overshoots the maximum of @xmath53 .
soon thereafter , the potential is well - approximated by @xmath54 , and the solution corresponds to @xmath55 in the matter frame .
thus , the universe evolves to late - time power - law inflation , with observational consequences similar to dark energy with equation - of - state parameter @xmath56 .
_ 3 . future singularity . _ for @xmath57 , @xmath19 does not reach the maximum of its potential and rolls back down to @xmath58 .
this yields a future curvature singularity .
what about including matter ?
as can be seen from ( [ confscalar ] ) , the major difference here is that the equation - of - motion for @xmath19 in the einstein frame has a new term .
furthermore , since the matter density is much greater than @xmath59 for @xmath60gyr , this term is very large and greatly affects the evolution of @xmath19 .
the exception is when the matter content is radiation alone ( @xmath61 ) , in which case it decouples from the @xmath19 equation due to conformal invariance . despite this complication ,
it is possible to show that the three possible cosmic futures identified in the vacuum case remain in the presence of matter . thus far
, the dimensionful parameter @xmath26 is unspecified . by choosing @xmath62ev ,
the corrections to the standard cosmology only become important at the present epoch , explaining the observed acceleration of the universe without recourse to dark energy .
clearly the choice of correction to the gravitational action can be generalized .
terms of the form @xmath63 , with @xmath64 , lead to similar late - time self acceleration , which can easily accommodate current observational bounds on the equation of state parameter . now ,
as i mentioned in the introduction , any modification of the einstein - hilbert action must , of course , be consistent with the classic solar system tests of gravity theory , as well as numerous other astrophysical dynamical tests .
we have chosen the coupling constant @xmath26 to be very small , but we have also introduced a new light degree of freedom .
as shown by chiba @xcite , the simple model above is equivalent to a brans - dicke theory with @xmath65 in the approximation where the potential was neglected , and would therefore be inconsistent with experiment @xcite ( although see @xcite for suggestions that the conformally transformed theory may not be the correct way to analyze deviations from gr ) . to construct a realistic @xmath0 model
requires a more complicated function , with more than one adjustable parameter in order to fit the cosmological data and satisfy solar system bounds .
examples can be found in @xcite .
it is natural to consider generalizing the action of @xcite to include other curvature invariants .
there are , of course , any number of terms that one could consider , but for simplicity , focus on those invariants of lowest mass dimension that are also parity - conserving @xmath66 we consider actions of the form @xmath67 + \int d^4 x\ , \sqrt{-g}\ , { \cal l}_m \ , \label{genaction}\ ] ] where @xmath68 is a general function describing deviations from general relativity .
it is convenient to define @xmath69 in terms of which the equations of motion are @xmath70 + \box(f_p\,r_{\mu\nu})\nonumber\\ & + & g_{\mu\nu}\,\nabla_\alpha\nabla_\beta(f_p\,r^{\alpha\beta } ) -4\nabla_\alpha\nabla_\beta[f_q\,r^\alpha{}_{(\mu\nu)}{}^\beta ] = 8\pi g\,t_{\mu\nu}\ .\label{equaz}\end{aligned}\ ] ] it is straightforward to show that actions of the form ( [ genaction ] ) generically admit a maximally - symmetric solution : @xmath71 a non - zero constant .
however , an equally generic feature of such models is that this de sitter solution is unstable . in the cdtt model
the instability is to an accelerating power - law attractor .
this is a possibility that we will also see in many of the more general models under consideration here .
since we are interested in adding terms to the action that explicitly forbid flat space as a solution , i will , in a similar way as in @xcite , consider inverse powers of the above invariants and , for simplicity , specialize to a class of actions with @xmath72 where @xmath73 is a positive integer ( taken to be unity ) , @xmath26 has dimensions of mass and @xmath74 , @xmath75 and @xmath76 are dimensionless constants .
in fact , for general @xmath73 the qualitative features of the system are as for @xmath77@xcite . for the purposes of this short talk , i will focus on a specific example - actions containing modifications involving only @xmath78 , with the prototype being @xmath79 , with @xmath80 a parameter with dimensions of mass .
it is easy to see that there is a constant curvature vacuum solution to this action given by @xmath81 .
however , we would like to investigate other cosmological solutions and analyze their stability . from ( [ equaz ] ) , with the flat cosmological ansatz , the analogue of the friedmann equation becomes 3h^2&- & = 0 .
[ riccieqn ] asymptotic analysis of this equation ( substituting in a power - law ansatz and taking the late - time limit ) yields two late - time attractors with powers @xmath82 and @xmath83 . however , in order to obtain a late - time accelerating solution ( @xmath84 ) , it is necessary to give accelerating initial conditions ( @xmath85 ) , otherwise the system is in the basin of attraction of the non - accelerating attractor at @xmath86 ( this type of behavior is generic in some other modified gravity theories @xcite).while i ve given a simple example here , cosmologically viable models are described in @xcite . what about the other constraints on these models ?
it has been shown @xcite that solar system constraints , of the type i have described for @xmath0 models , can be evaded by these more general models whenever the constant @xmath76 is nonzero . roughly speaking ,
this is because the schwarzschild solution , which governs the solar system , has vanishing @xmath1 and @xmath87 , but non - vanishing @xmath88 .
more serious is the issue of ghosts and superluminal propagation .
it has been shown @xcite that a necessary but not sufficient condition that the action be ghost - free is that @xmath89 , so that there are no fourth derivatives in the linearised field equations . what remained was the possibility that the second derivatives might have the wrong signs , and also might allow superluminal propagation at some time in a particular cosmological background .
it has recently been shown that in a frw background with matter , the theories are ghost - free , but contain superluminally propagating scalar or tensor modes over a wide range of parameter space @xcite .
it is certainly necessary to be ghost - free .
whether the presence of superluminally propagating modes is a fatal blow to the theories remains to be seen .
given the immense challenge posed by the accelerating universe , it is important to explore every option to explain the underlying physics
. modifying gravity may be one of the more radical proposals , but it is not one without precedence as an explanation for unusual physics .
however , it is an approach that is tightly constrained both by observation and theoretical obstacles . in the brief time and space
allowed , i have tried to give a flavor of some attempts to modify gr to account for cosmic acceleration without dark energy .
i have focused on two of the directions in which i have been involved and have chosen to present simple examples of the models , which clearly demonstrate not only the cosmological effects , but also how constraints from solar system tests and theoretical consistency apply .
there are a number of other proposals for modified gravity and , while i have had neither time nor space to devote to them here , others have discussed some of them in detail at this meeting .
there is much work ahead , with significant current effort , my own included , devoted to how one might distinguish between modified gravity , dark energy and a cosmological constant as competing explanations for cosmic acceleration .
i would like to thank the organizers of the q2c conference , and in particular slava turyshev , for their hard work and dedication in running such a stimulating meeting .
i would also like to thank my many coauthors on the work discussed here for such enjoyable and productive collaborations , and for allowing me to reproduce parts of our work in this article .
this work was supported in part by the nsf under grant phy-0354990 , by research corporation , and by funds provided by syracuse university 0 g. r. dvali , g. gabadadze and m. porrati , phys .
b * 485 * , 208 ( 2000 ) [ arxiv : hep - th/0005016 ] . c. deffayet , phys .
b * 502 * , 199 ( 2001 ) [ arxiv : hep - th/0010186 ] . c. deffayet , g. r. dvali and g. gabadadze , phys .
d * 65 * , 044023 ( 2002 ) [ arxiv : astro - ph/0105068 ] .
k. freese and m. lewis , phys .
b * 540 * , 1 ( 2002 ) [ arxiv : astro - ph/0201229 ] .
g. dvali and m. s. turner , arxiv : astro - ph/0301510 . s. m. carroll , v. duvvuri , m. trodden and m. s. turner , phys . rev .
d * 70 * , 043528 ( 2004 ) [ arxiv : astro - ph/0306438 ] .
s. capozziello , s. carloni and a. troisi , arxiv : astro - ph/0303041 .
d. n. vollick , phys .
d * 68 * , 063510 ( 2003 ) [ arxiv : astro - ph/0306630 ] .
e. e. flanagan , phys .
lett . * 92 * , 071101 ( 2004 ) [ arxiv : astro - ph/0308111 ] .
e. e. flanagan , class .
* 21 * , 417 ( 2003 ) [ arxiv : gr - qc/0309015 ] .
d. n. vollick , `` on the viability of the palatini form of 1/r gravity , '' class .
* 21 * , 3813 ( 2004 ) [ arxiv : gr - qc/0312041 ] .
m. e. soussa and r. p. woodard , gen .
grav . * 36 * , 855 ( 2004 ) [ arxiv : astro - ph/0308114 ] .
s. nojiri and s. d. odintsov , gen .
grav . * 36 * , 1765 ( 2004 ) [ arxiv : hep - th/0308176 ] .
s. m. carroll , a. de felice , v. duvvuri , d. a. easson , m. trodden and m. s. turner , phys .
d * 71 * , 063513 ( 2005 ) [ arxiv : astro - ph/0410031 ] .
n. arkani - hamed , h. c. cheng , m. a. luty and s. mukohyama , jhep * 0405 * , 074 ( 2004 ) [ arxiv : hep - th/0312099 ] . g. gabadadze and m. shifman , phys .
d * 69 * , 124032 ( 2004 ) [ arxiv : hep - th/0312289 ] .
j. w. moffat , arxiv : astro - ph/0403266 .
t. clifton , d. f. mota and j. d. barrow , mon . not .
soc . * 358 * , 601 ( 2005 ) [ arxiv : gr - qc/0406001 ] .
s. m. carroll , i. sawicki , a. silvestri and m. trodden , arxiv : astro - ph/0607458 .
j. d. barrow and a. c. ottewill , j. phys .
a * 16 * , 2757 ( 1983 ) .
j. d. barrow and s. cotsakis , phys .
b * 214 * , 515 ( 1988 ) .
j. d. barrow and s. cotsakis , phys .
b * 258 * , 299 ( 1991 ) .
g. magnano and l. m. sokolowski , phys .
d * 50 * , 5039 ( 1994 ) [ arxiv : gr - qc/9312008 ] .
a.dobado and a.l .
maroto , phys .
d * 52 * , 1895 ( 1995 ) .
h. j. schmidt , astron .
* 311 * , 165 ( 1990 ) [ arxiv : gr - qc/0109004 ] .
g. magnano and l. m. sokolowski , phys .
d * 50 * , 5039 ( 1994 ) [ arxiv : gr - qc/9312008 ] .
t. chiba , phys .
b * 575 * , 1 ( 2003 ) [ arxiv : astro - ph/0307338 ] . b. bertotti , l. iess and p. tortora , nature * 425 * , 374 ( 2003 ) .
v. faraoni , arxiv : gr - qc/0607016 .
s. capozziello and a. troisi , phys .
d * 72 * , 044022 ( 2005 ) [ arxiv : astro - ph/0507545 ] .
s. capozziello , a. stabile and a. troisi , higher order gravity compatible with experimental constraints on eddington arxiv : gr - qc/0603071 .
p. zhang , phys .
d * 73 * , 123504 ( 2006 ) [ arxiv : astro - ph/0511218 ] .
d. a. easson , f. p. schuller , m. trodden and m. n. r. wohlfarth , phys .
d * 72 * , 043504 ( 2005 ) [ arxiv : astro - ph/0506392 ] .
o. mena , j. santiago and j. weller , phys .
* 96 * , 041103 ( 2006 ) [ arxiv : astro - ph/0510453 ] .
i. navarro and k. van acoleyen , phys .
b * 622 * , 1 ( 2005 ) [ arxiv : gr - qc/0506096 ] .
t. chiba , jcap * 0503 * , 008 ( 2005 ) [ arxiv : gr - qc/0502070 ] .
i. navarro and k. van acoleyen , consistent long distance modification of gravity from inverse powers of the jcap * 0603 * , 008 ( 2006 ) [ arxiv : gr - qc/0511045 ] . a. de felice , m. hindmarsh and m. trodden , arxiv : astro - ph/0604154 .
g. calcagni , b. de carlos and a. de felice , arxiv : hep - th/0604201 . | i briefly discuss some attempts to construct a consistent modification to general relativity ( gr ) that might explain the observed late - time acceleration of the universe and provide an alternative to dark energy .
i mention the issues facing extensions to gr , illustrate these with two specific examples , and discuss the resulting observational and theoretical obstacles .
this article comprises an invited talk at the nasa workshop _ from quantum to cosmos : fundamental physics research in space_. |
low - energy dynamics in one - dimensional ( 1d ) quantum magnets is a long - standing problem and nuclear magnetic resonance ( nmr ) findings are often enlightening in this context .
it was a fine collaboration that the nmr relaxation rates @xmath2 and @xmath3 ( ref . ) of the spin-@xmath4 antiferromagnet sr@xmath5cuo@xmath6 illuminated multiplicative logarithmic corrections to the dynamic susceptibility @xcite in critical spin chains .
@xmath0 measurements on the spin-@xmath7 antiferromagnet ( ch@xmath6)@xmath8nmncl@xmath6 were a pioneering study in an attempt to detect the long - time diffusive spin dynamics in one dimension .
@xcite the diffusive contribution to @xmath2 was observed in the spin-@xmath4 quantum limit @xcite and for spin - gapped antiferromagnets @xcite as well .
antiferromagnets , whether critical or gapped , are thus vigorously studied , while little is known about ferrimagnetic , as well as ferromagnetic , dynamics . recently
numerous 1d ferrimagnets have been synthesized in an effort to design molecule - based ferromagnets .
the static properties of various heterospin chains were correspondingly calculated , @xcite but few effort has been devoted to exploring their dynamic features .
recent nmr observations of 1d ferrimagnets @xcite have thus been motivated and have indeed stimulated the theoretical interest in them .
the most pioneering @xmath0 measurements , @xcite performed on the metal - radical hybrid compound mn(c@xmath9h@xmath5o@xmath5f@xmath10)@xmath5c@xmath11h@xmath12n@xmath5o@xmath5 , were interpreted in terms of solitonic excitations at low temperatures and based on the spin - diffusion model at high temperatures . however , this material has rather large exchange interactions and its magnetic susceptibility ( @xmath13 ) times temperature ( @xmath14 ) exhibits no ferrimagnetic minimum at a measurable temperature .
@xcite the nonnegligible single - ion anisotropy and the spread magnetic moment over the radical may also blur the intrinsic ferrimagnetic features .
the family of manganeseporphyrin - based ferrimagnets , whose exchange interactions are somewhat smaller , was also observed through nmr , @xcite but the complicated crystalline structure restricted the analysis to solving the magnetic configuration . @xmath15
nmr findings @xcite on the bimetallic chain compound nicu(c@xmath16h@xmath10n@xmath5o@xmath10)(h@xmath5o)@xmath6@xmath172h@xmath5o implied a spin - wave relaxation scenario , but the averaging effect over the numerous protons masks the low - temperature dynamics .
trimeric chains in ca@xmath6cu@xmath6(po@xmath8)@xmath8 .
the strongly coupled cu@xmath18 trimer consists of a central square planar cu(1 ) ion ( black circle ) and two pyramidal cu(2 ) ions ( gray circles).,width=226 ] .,width=317 ] in such circumstances , we perform an nmr study on the homometallic chain compound ca@xmath6cu@xmath6(po@xmath8)@xmath8 ( fig . [
f : illust ] ) , which is a novel ferrimagnet of topological origin .
@xcite this collaboration aims at verifying a newly developed modified spin - wave nuclear magnetic relaxation theory . @xcite
two of the present authors have recently demonstrated that @xmath2 in 1d heterometallic ferrimagnets may significantly be enhanced by exchange interactions , in an attempt to interpret the novel nmr observations @xcite for nicu(c@xmath16h@xmath10n@xmath5o@xmath10)(h@xmath5o)@xmath6@xmath172h@xmath5o , which are hardly understandable within the usual raman relaxation scheme . however , all the findings @xcite were far from conclusive to a possibility of more than two magnons mediating the proton spin relaxation . since only the @xmath15 nuclei were practically available as probes in nicu(c@xmath16h@xmath10n@xmath5o@xmath10)(h@xmath5o)@xmath6@xmath172h@xmath5o , the crystal water of nuisance and the resultant ill - behaving spin - echo recovery curve with increasing field at low temperatures restricted the @xmath0 analysis to rather high temperatures @xmath19 .
ca@xmath6cu@xmath6(po@xmath8)@xmath8 contains p atoms as efficient probes . the well - isolated cu@xmath18 chains with ca columns in between , no single - ion anisotropy of the cu@xmath18 spins , and the ferrimagnetic minimum of @xmath20 at a moderate temperature [ see fig .
[ f : chi](b ) ] furthermore guarantee this material to be a fine stage of 1d ferrimagnetic dynamics .
we describe ca@xmath6cu@xmath6(po@xmath8)@xmath8 by the hamiltonian , @xmath21 , \label{e : h}\end{aligned}\ ] ] where each cu(1 ) ion is antiferromagnetically coupled to four cu(2 ) ions in an applied field ( @xmath22 ) ( see fig .
[ f : illust ] ) .
we set @xmath23 and @xmath24 equal to @xmath25 and @xmath26 , respectively .
@xcite employing the holstein - primakoff transformation , @xcite we expand the hamiltonian with respect to @xmath27 as @xmath28 where @xmath29 contains the @xmath30 terms .
@xmath31 describes linear spin - wave excitations and is diagonalized in the momentum space as @xmath32 where @xmath33 creates a spin wave of ferromagnetic ( @xmath34 ) or antiferromagnetic ( @xmath35 ) aspect , whose excitation energy is given by @xmath36 with @xmath37 [ see fig .
[ f : chi](a ) ] .
@xmath38 gives two - body interactions and makes a crucial contribution to nuclear spin - lattice relaxation .
the @xmath39 terms are neglected in the following .
the dispersive branches @xmath40 and @xmath41 are reminiscent of the dual excitations in alternating - spin chains , @xcite whereas the flat band @xmath42 , describing intratrimer excitations , is peculiar to the present system .
our way @xcite of modifying the conventional spin - wave theory is distinct from the original idea proposed by takahashi @xcite and hirsch _
et al._. @xcite their way of suppressing the divergent sublattice magnetizations consists of diagonalizing the hamiltonian together with a lagrange multiplier subject to zero staggered magnetization .
the thus - obtained energy spectrum depends on temperature and fails , for instance , to reproduce the schottky - peaked specific heat .
seeking after better thermodynamics , we diagonalize the bare hamiltonian and then minimize the free energy with a lagrange multiplier subject to zero staggered magnetization .
@xcite the thus - calculated @xmath13 is in good agreement with observations [ fig .
[ f : chi](b ) ] .
the hyperfine interaction between a @xmath1 nucleus and cu@xmath18 spins consists of isotropic fermi contact and anisotropic dipolar coupling and is defined as @xmath43 @xmath38 and @xmath44 are both much smaller than @xmath31 and may be regarded as perturbations to the linear spin - wave system .
when we calculate up to second order in @xmath45 , the probability of a nuclear spin being scattered from the state of @xmath46 to that of @xmath47 is given by @xmath48 where @xmath49 and @xmath50 denote the initial and final states of the unperturbed electronic - nuclear spin system .
then we obtain @xmath51 . considering the significant difference between the nuclear and electronic energy scales at moderate fields and assuming the fourier components of the coupling tensors to have little momentum dependence as @xmath52
, the raman and three - magnon relaxation rates read @xmath53 , \label{e : t1(3)}\end{aligned}\ ] ] where @xmath54 is the thermal distribution function of modified spin waves , while @xmath55 and @xmath56 are determined through @xmath57 , @xmath58 , @xmath59 , and @xmath60 .
@xmath61 and @xmath62 are diagrammatically represented in fig .
[ f : diagram ] . within the first - order mechanism containing a nuclear spin in direct contact with spin waves via the hyperfine interaction ,
any multi - magnon relaxation rate is much smaller than the raman one .
however , some of multi - magnon processes make a significant contribution to @xmath2 through the second - order mechanism , where a nuclear spin flips with the help of _ virtual spin waves _ which are then scattered thermally via the four - magnon exchange interaction . we consider the leading second - order relaxation , that is , _ exchange - scattering - induced three - magnon processes _ , as well as the first - order relaxation .
second - order single - magnon and raman relaxation processes , containing three and two virtual magnons , respectively , are much more accidental due to the momentum conservation and much less contributive due to the magnon series damping . as for four - magnon processes ,
the first - order relaxation is nonexistent to begin with , whereas the second - order one originates in the six - magnon exchange interaction and contains two virtual magnons .
thus and thus , all other higher - order processes have no significant effect on the relaxation scenario .
nmr spectra measured at @xmath63 .
they can be decomposed into typical powder patterns [ the insets ( a ) and ( b ) ] , originating in the p(1 ) ( dotted lines ) and p(2 ) ( broken lines ) sites , and the nmr shift for the latter is plotted as a function of the susceptibility at @xmath64 [ the inset ( c)].,width=317 ]
now we present @xmath1 nmr spectra for powder samples of ca@xmath6cu@xmath6(po@xmath8)@xmath8 in fig .
[ f : spectra ] .
the crystallographically inequivalent p sites , labeled p(1 ) and p(2 ) , give two distinct lines , each of which has a characteristic shape of the anisotropic powder pattern , as is revealed in the insets ( a ) and ( b ) .
both lines broaden and shift to lower field with decreasing temperature , implying that contact as well as dipolar terms exist in the hyperfine field on the p nuclei .
the nmr shifts are thus anisotropic and their parallel ( @xmath65 ) and perpendicular ( @xmath66 ) components for the p(2 ) sites are plotted as functions of the susceptibility at @xmath64 in the inset ( c ) . if we roughly estimate the hyperfine coupling between each p(2 ) nucleus and the nearest - neighbor cu@xmath18 ion through the relations @xmath67 and @xmath68 , where @xmath69 is the avogadro number , we find that @xmath70 and @xmath71 . in order to reveal the low - frequency spin dynamics , we measure @xmath0 for the p(2 ) sites through the saturation - recovery method . as functions of temperature . @xmath72 and @xmath73 are also plotted in ( b).,width=325 ] as functions of @xmath74 .
dotted lines are guides to the probable exponential behavior @xmath75 at each field .
( b ) the thus - extracted @xmath76 as a function of field .
the two slopes , @xmath77 and @xmath78 , are shown for reference.,width=325 ] temperature dependences of @xmath2 are shown in fig .
[ f : t1 t ] . in each cu@xmath18 trimer unit , the two spins @xmath79 and
@xmath80 are almost equidistant from the nearby p(2 ) atom , whereas the other one @xmath81 is much more distant from that .
@xcite therefore , the isotropic coupling constants are taken as @xmath82 and @xmath83 . since
dipolar interactions are also sensitive to the location of correlating moments , the anisotropic coupling constants may be taken similarly as @xmath84 and @xmath85
. then we set @xmath86 and @xmath87 equal to @xmath88 and @xmath89 , respectively , which are both consistent well with the experimental findings @xmath70 and @xmath71 .
considering that recent electron - spin - resonance measurements of this compound @xcite have yield temperature - dependent and anisotropic @xmath90 values ( @xmath91 ) , the @xmath92 value may be closer to the theoretical parametrization .
we are further convinced of the coupling constants employed finding the nearest cu(1)-p(2 ) distance to be about @xmath93 , which is in excellent agreement with the crystalline structure .
@xcite the thus - calculated @xmath94 reproduces the observations pretty well .
the exchange - scattering - enhanced three - magnon relaxation rate generally grows into a major contribution to @xmath2 with increasing temperature and decreasing field . while both @xmath72 and @xmath73 exhibit an exponential behavior at low temperatures , their activation energies , referred to as @xmath95 and @xmath96
, respectively , look different . at moderately low temperatures and weak fields , @xmath97 , eq .
( [ e : t1(2 ) ] ) reads @xmath98 where @xmath99 is the modified bessel function of the second kind and behaves as @xmath100 .
thus we learn that @xmath101 .
equation ( [ e : t1(3 ) ] ) is much less analyzable , but fig .
[ f : t1t](b ) claims that @xmath102 .
figure [ f : t1gap ] brings the next leading exponential behavior @xmath75 to light , because eq . (
[ e : t1(2)ap ] ) is valid for @xmath103 .
@xmath76 looks like @xmath104 or more , rather than @xmath105 .
the two - magnon - mediated nuclear spin relaxation is mainly given by @xmath106 , where a ferromagnetic spin wave of energy @xmath107 is created and that of energy @xmath108 is destructed , while the three - magnon - mediated one by @xmath109 , where two net spin waves of energy @xmath110 and @xmath111 are created and that of energy @xmath112 is destructed .
therefore , raman processes are activated by the energy @xmath113 , which is consistent with eq .
( [ e : t1(2)ap ] ) , whereas three - magnon ones roughly by the energy @xmath114 , which may look somewhat larger due to the complicated field dependence lying in the second - order mechanism . as functions of field at @xmath115 ( a ) and @xmath116 ( b ) , where @xmath72 and @xmath73 are also plotted .
the observations are replotted as functions of @xmath117 and fitted to an expression @xmath118 ( c).,width=325 ] field dependences of @xmath2 at higher temperatures more impress on us the significance of three - magnon processes .
figures [ f : t1h](a ) and [ f : t1h](b ) show that the accelerated relaxation with decreasing field can never be explained by the raman scheme but should be attributed to exchange - scattering - enhanced three - magnon processes .
the spin - diffusion model may be mentioned in this context .
diffusion - dominated 1d spin dynamics gives @xmath2 of the form @xmath119 , @xcite where the first and second terms come from transverse and longitudinal spin fluctuations , respectively , and are both positive .
figure [ f : t1h](c ) shows that the present observations fitted to the diffusive law result in negative @xmath120 .
we do not exclude a possibility of diffusive dynamics appearing in 1d ferrimagnets as well , but a distinct field dependence of the second - order relaxation mechanism masks such a moderate field effect in the present case .
it was in agvp@xmath5s@xmath10 rather than in the most familiar haldane - gap antiferromagnet ni(c@xmath5h@xmath121n@xmath5)@xmath5no@xmath5clo@xmath8 , whose excitation spectrum drastically varies with increasing field , that integral - spin diffusive correlations were observed .
@xcite exchange - scattering - induced three - magnon processes are sensitive to an excitation gap and their contribution to @xmath2 is strongly suppressed , for example , by slight magnetic anisotropy .
there are indeed some indications of spin diffusion @xcite in the ferrimagnetic chain compound mn(c@xmath9h@xmath5o@xmath5f@xmath10)@xmath5c@xmath11h@xmath12n@xmath5o@xmath5 with nonnegligible single - ion anisotropy .
we have performed nmr measurements on the topological ferrimagnet ca@xmath6cu@xmath6(po@xmath8)@xmath8 and have confirmed a novel scenario for 1d spin dynamics@xmath122_multimagnon - mediated nuclear spin relaxation _ , by showing the parametrization to be crystallographically convincing , revealing that the relaxation is activated by twice the gap rather than the gap itself at low temperatures , while it is remarkably accelerated with decreasing field , and pointing out the irrelevance of the spin - diffusion model . indeed pioneering @xmath0 observations on the layered ferromagnet crcl@xmath6 ( ref . ) and the coupled - chain antiferromagnet csmncl@xmath1232h@xmath5o ( ref . ) claimed to have detected three - magnon processes , but they were both , in some sense , classical findings under the 3d long - range order .
@xcite our findings are literally 1d quantum spin relaxation beyond the raman mechanism , which were obtained through an elaborately modified spin - wave theory . @xcite the bond - alternating homometallic chain compound cu(c@xmath9clh@xmath8n)@xmath5(n@xmath6)@xmath5 ( ref . )
is another anisotropy - free ferrimagnet of topological origin @xcite and therefore nmr measurements on it are highly encouraged . since 1d ferromagnets , which can be regarded as low - energy sectors of 1d ferrimagnets ,
may also play this fascinating scenario , more understanding will come with further experiments , for instance , on the spin-@xmath4 ferromagnetic chain compound ( ch@xmath6)@xmath8ncucl@xmath6 .
@xcite the quasi - one - dimensional mixed - spin ferromagnet @xmath124 ( ref . ) is also highly interesting in this context , whose low - energy spectrum consists of two dispersive ferromagnetic excitation branches | the nuclear spin - lattice relaxation time @xmath0 of @xmath1 nuclei in the title compound is measured for the first time and interpreted in terms of a modified spin - wave theory .
we establish a novel scenario for one - dimensional ferrimagnetic spin dynamics it is _ three - magnon processes enhanced by exchange scattering _ , rather than raman processes , that make the major contribution to @xmath2 . |
the possibility to predict future states of a system stands at the foundations of scientific knowledge with an obvious relevance both from a conceptual and applicative point of view .
the perfect knowledge of the evolution law of a system may induce the conclusion that this aim could be attained .
this classical deterministic point of view was claimed by laplace @xcite : once the evolution laws of the system are known , the state at a certain time @xmath0 completely determines the subsequent states for every time @xmath1 .
however it is well established now that in some systems , full predictability can not be accomplished in practice because of the unavoidable uncertainty in the initial conditions .
indeed , as already stated by poincar , long - time predictions are reliable only when the evolution law does not amplify the initial uncertainty too rapidly .
therefore , from the point of view of predictability , we need to know how an error on the initial state of the system grows in time . in systems with great sensitive dependence on initial conditions ( deterministic chaotic systems ) errors
grows exponentially fast in time , limiting the ability to predict the future states .
a branch of the theory of dynamical systems has been developed with the aim of formalizing and characterizing the sensitivity to initial conditions .
the lyapunov exponent and the kolmogorov - sinai entropy are the two main indicators for measuring the rate of error growth and information production during a deterministic system evolution .
a complementary approach has been developed in the context of information theory , data compression and algorithmic complexity theory and it is rather clear that the latter point of view is closely related to the dynamical systems one .
if a system is chaotic , then its predictability is limited up to a time which is related to the first lyapunov exponent , and the time sequence by which we encode one of its chaotic trajectories can not be compressed by an arbitrary factor , i.e. is algorithmically complex . on the contrary
, the coding of a regular trajectory can be easily compressed ( e.g. , for a periodic trajectory it is sufficient to have the sequence for a period ) so it is `` simple '' . in this paper
we will discuss how unpredictability and algorithmic complexity are closely related and how information and chaos theory complete each other in giving a general understanding of complexity in dynamical processes .
in particular , we shall consider the extension of this approach , nowadays well established in the context of low dimensional systems and for asymptotic regimes , to high dimensional systems with attention to situations far from asymptotic ( i.e. finite time and finite observational resolution ) @xcite .
the characteristic lyapunov exponents are somehow an extension of the linear stability analysis to the case of aperiodic motions . roughly speaking , they measure the typical rate of exponential divergence of nearby trajectories and , thus , contain information on the growing rate of a very small error on the initial state of a system . consider a dynamical system with an evolution law given , e.g. , by the differential equation @xmath2 we assume that @xmath3 is smooth enough that the evolution is well - defined for time intervals of arbitrary extension , and that the motion occurs in a bounded region of the phase space .
we intend to study the separation between two trajectories , @xmath4 and @xmath5 , starting from two close initial conditions , @xmath6 and @xmath7 , respectively . as long as the difference between the trajectories , @xmath8 , remains small ( infinitesimal , strictly speaking )
, it can be regarded as a vector , @xmath9 , in the tangent space .
the time evolution of @xmath9 is given by the linearized differential equations : @xmath10 under rather general hypothesis , oseledec @xcite proved that for almost all initial conditions @xmath6 , there exists an orthonormal basis @xmath11 in the tangent space such that , for large times , @xmath12 where the coefficients @xmath13 depend on @xmath14 .
the exponents @xmath15 are called _ characteristic lyapunov exponents _ ( les ) .
if the dynamical system has an ergodic invariant measure , the spectrum of les @xmath16 does not depend on the initial condition , except for a set of measure zero with respect to the natural invariant measure .
equation ( [ eq:1 - 5 ] ) describes how a @xmath17-dimensional spherical region of the phase space , with radius @xmath18 centered in @xmath19 , deforms , with time , into an ellipsoid of semi - axes @xmath20 , directed along the @xmath21 vectors .
furthermore , for a generic small perturbation @xmath22 , the distance between the reference and the perturbed trajectory behaves as @xmath23.\ ] ] if @xmath24 we have a rapid ( exponential ) amplification of an error on the initial condition . in such a case ,
the system is chaotic and , _ de facto _ , unpredictable on the long times . indeed ,
if the initial error amounts to @xmath25 , and we purpose to predict the states of the system with a certain tolerance @xmath26 ( not too large ) , then the prediction is reliable just up to a _ predictability time _ given by @xmath27 this equation shows that @xmath28 is basically determined by the largest lyapunov exponent , since its dependence on @xmath29 and @xmath26 is logarithmically weak .
because of its preeminent role , @xmath30 is often referred as `` the lyapunov exponent '' , and denoted by @xmath31 . in experimental investigations of physical processes ,
the access to a system occurs only through a measuring device which produces a time record of a certain observable , i.e. a sequence of data . in this
regard a system , whether or not chaotic , generates messages and may be regarded as a source of information whose properties can be analysed through the tools of information theory .
the characterization of the information contained in a sequence can be approached in two very different frameworks .
the first considers a specific message ( sequence ) as belonging to the ensemble of all the messages that can be emitted by a source , and defines an average information content by means of the average compressibility properties of the ensemble @xcite .
the second considers the problem of characterizing the universal compressibility ( i.e. ensemble independent ) of a specific sequence and concerns the theory of algorithmic complexity and algorithmic information theory @xcite . for the sake of self - consistency we briefly recall the concepts and ideas about the shannon entropy@xcite , that is the basis of whole information theory
consider a source that can output @xmath32 different symbols ; denote with @xmath33 the symbol emitted by the source at time @xmath34 and with @xmath35 the probability that a given word @xmath36 , of length @xmath37 , is emitted @xmath38 .
we assume that the source is stationary , so that , for the sequences @xmath39 , the time translation invariance holds : @xmath40 .
we introduce the @xmath37-block entropies @xmath41 for stationary sources the limit @xmath42 exists and defines the shannon entropy @xmath43 which quantifies the richness ( or `` complexity '' ) of the source emitting the sequence .
this can be precisely expressed by the first theorem of shannon - mcmillan @xcite that applies to stationary ergodic sources : the ensemble of @xmath37-long subsequences , when @xmath37 is large enough , can be partitioned in two classes , @xmath44 and @xmath45 such that all the words @xmath46 have the same probability @xmath47 and @xmath48 the meaning of this theorem is the following .
an @xmath32-states process admits , in principle , @xmath49 possible sequences of length @xmath37 .
however the number of typical sequences , @xmath50 , effectively observable ( i.e. those belonging to @xmath44 ) is @xmath51 note that @xmath52 if @xmath53 .
the entropy per symbol , @xmath43 , is a property of the source .
however , because of the ergodicity @xmath43 can be obtained by analyzing just one single sequence in the ensemble of the typical ones , and it can also be viewed as a property of each typical sequence . in information theory , expression ( [ eq : wordstypical ] ) is somehow the equivalent of the boltzmann equation in statistical thermodynamics : @xmath54 , being @xmath55 the number of possible microscopic configurations and @xmath56 the thermodynamic entropy , this justifies the name `` entropy '' for @xmath43 .
the relevance of the shannon entropy in information theory is given by the fact that @xmath43 sets the maximum compression rate of a sequence @xmath57 .
indeed a theorem of shannon states that , if the length @xmath58 of a sequence is large enough , there exists no other sequence ( always using @xmath32 symbols ) , from which it is possible to reconstruct the original one , whose length is smaller than @xmath59 @xcite . in other words , @xmath60 represents the maximum allowed compression rate . the relation between shannon entropy and data compression problems
is well illustrated by considering the optimal coding ( shannon - fano ) to map @xmath61 objects ( e.g. the @xmath37-words @xmath62 ) into sequences of binary digits @xmath63 @xcite . denoting with @xmath64 the binary length of the sequence specifying @xmath62 , we have @xmath65 i.e. , in a good coding , the mean length of a @xmath37-word is equal to @xmath37 times the shannon entropy , apart from a multiplicative factor , since in the definition ( [ eq : shannon ] ) of @xmath43 we used the natural logarithm and here we want to work with a two symbol code . after the introduction of the shannon entropy we can easily define the kolmogorov - sinai entropy which is the analogous measure of complexity applied to dynamical systems .
consider a trajectory , @xmath4 , generated by a deterministic system , sampled at the times @xmath66 , with @xmath67 .
perform a finite partition @xmath68 of the phase space , with the finite number of symbols @xmath69 enumerating the cells of the partition . the time - discretized trajectory @xmath70 determines a sequence @xmath71 , whose meaning is clear : at the time @xmath72 the trajectory is in the cell labeled by @xmath73 .
to each subsequence of length @xmath74 one can associate a word of length @xmath37 : @xmath75 .
if the system is ergodic , as we suppose , from the frequencies of the words one obtains the probabilities by which the block entropies @xmath76 are calculated : @xmath77 the probabilities @xmath78 , computed by the frequencies of @xmath79 along a trajectory , are essentially dependent on the stationary measure selected by the trajectory .
the entropy per unit time of the trajectory with respect to the partition @xmath68 , @xmath80 , is defined as follows : @xmath81 notice that , for the deterministic systems we are considering , the entropy per unit time does not depend on the sampling time @xmath82 @xcite .
the ks - entropy ( @xmath83 ) , by definition , is the supremum of @xmath84 over all possible finite partitions @xcite @xmath85 the extremal character of @xmath83 makes every computation based on the definition ( [ eq : ks ] ) , impossible in the majority of practical cases . in this respect
, a useful tool would be the kolmogorov - sinai theorem , through which one is granted that @xmath86 if @xmath87 is a generating partition .
a partition is said to be generating if every infinite sequence @xmath88 corresponds to a single initial point .
however the difficulty now is that , with the exception of very simple cases , we do not know how to construct a generating partition .
we only know that , according to the krieger theorem @xcite , there exists a generating partition with @xmath89 elements such that @xmath90 .
then , a more tractable way to define @xmath83 is based upon considering the partition @xmath91 made up by a grid of cubic cells of edge @xmath18 , from which one has @xmath92 we expect that @xmath93 becomes independent of @xmath18 when @xmath94 is so fine to be `` contained '' in a generating partition . for discrete time maps
what has been exposed above is still valid , with @xmath95 ( however , krieger s theorem only applies to invertible maps ) .
the important point to note is that , for a truly stochastic ( i.e. non - deterministic ) system , with continuous states , @xmath96 is not bounded and @xmath97 .
the shannon entropy establishes a limit on how efficiently the ensemble of messages emitted by a source can be coded .
however , we may wonder about the compressibility properties of a single sequence with no reference to its belonging to an ensemble . that is to say
, we are looking for an universal characterization of its compressibility or , it is the same , an universal definition of its information content .
this problem can be addressed through the notion of _ algorithmic complexity _ , that concerns the difficulty in reproducing a given string of symbols .
everybody agrees that the binary digits sequence @xmath98 is , in some sense , more random than @xmath99 the notion of algorithmic complexity , independently introduced by kolmogorov @xcite , chaitin @xcite and solomonov @xcite , is a way to formalize the intuitive idea of randomness of a sequence .
consider , for instance , a binary digit sequence ( this does not constitute a limitation ) of length @xmath37 , @xmath100 , generated by a certain computer code on a given machine @xmath101 .
the algorithmic complexity ( or algorithmic information content ) @xmath102 of @xmath103 is the bit - length of the shortest computer program able to give @xmath103 and to stop afterward .
of course , such a length depends not only on the sequence but also on the machine . however , kolmogorov @xcite proved the existence of a universal computer , @xmath104 , able to perform the same computation that a program @xmath105 makes on @xmath106 , with a modification of @xmath105 that depends only on @xmath106 .
this implies that for all finite strings : @xmath107 where @xmath108 is the complexity with respect to the universal computer and @xmath109 depends only on the machine @xmath106 .
we can consider the algorithmic complexity with respect to a universal computer dropping the @xmath106-dependence in the symbol for the algorithmic complexity , @xmath110 .
the reason is that we are interested in the limit of very long sequences , @xmath111 , for which one defines the algorithmic complexity per unit symbol : @xmath112 that , because of ( [ eq : kolmocomplex ] ) , is an intrinsic quantity , i.e. independent of the machine . now coming back to the @xmath37-sequences ( [ eq : seq1 ] ) and ( [ eq : seq2 ] )
, it is obvious that the latter can be obtained with a minimal program of length @xmath113 and therefore when taking the limit @xmath111 in ( [ eq : acomplexity ] ) , one obtains @xmath114 . of course @xmath110
can not exceed @xmath37 , since the sequence can always be generated by a trivial program ( of bit length @xmath37 ) @xmath115 therefore , in the case of a very irregular sequence , e.g. , ( [ eq : seq1 ] ) , one expects @xmath116
( i.e. @xmath117 ) , and the sequence is named complex ( i.e. of non zero algorithmic complexity ) or random .
algorithmic complexity can not be computed , and the un - computability of @xmath110 may be understood in terms of gdel s incompleteness theorem @xcite . beyond the problem of whether or not @xmath110 is computable in a specific case , the concept of algorithmic complexity brings an important improvement to clarify the vague and intuitive notion of randomness . between the shannon entropy , @xmath43 , and the algorithmic complexity , there exists the straightforward relationship @xmath118 where @xmath119 , being @xmath120 the algorithmic complexity of the @xmath37-words , in the ensemble of sequences , @xmath62 , with a given distribution of probabilities , @xmath35 . therefore the expected complexity @xmath121 is asymptotically equal to the shannon entropy ( modulo the @xmath122 factor ) .
it is important to stress again that , apart from the numerical coincidence of the values of @xmath123 and @xmath124 , there is a conceptual difference between the information theory and the algorithmic complexity theory .
the shannon entropy essentially refers to the information content in a statistical sense , i.e. it refers to an ensemble of sequences generated by a certain source .
the algorithmic complexity defines the information content of an individual sequence @xcite .
the notion of algorithmic complexity can be also applied to the trajectories of a dynamical system .
this requires the introduction of finite open coverings of the phase space , the corresponding encoding of trajectories into symbolic sequences , and the searching of the supremum of the algorithmic complexity per symbol at varying the coverings @xcite .
brudno s and white s theorems @xcite state that the complexity @xmath125 for a trajectory starting from the point @xmath126 , is @xmath127 for almost all @xmath126 with respect to the natural invariant measure .
the factor @xmath122 stems again from the conversion between natural logarithms and bits .
this result indicates that the ks - entropy quantifies not only the richness of a dynamical system but also the difficulty of describing its typical sequences .
let us consider a @xmath128 chaotic map @xmath129 the transmission of the sequence @xmath130 , accepting only errors smaller than a tolerance @xmath26 , is carried out by using the following strategy
@xcite : 1 . transmit the rule ( [ eq : mappa ] ) : for this task one has to use a number of bits independent of the sequence length @xmath58 .
2 . specify the initial condition @xmath131 with a precision @xmath29 using a finite number of bits which is independent of @xmath58 .
3 . let the system evolve till the first time @xmath132 such that the distance between two trajectories , that was initially @xmath133 , equals @xmath26 and then specify again the new initial condition @xmath134 with precision @xmath29 .
4 . let the system evolve and repeat the procedure ( 2 - 3 ) , i.e. each time the error acceptance tolerance is reached specify the initial conditions , @xmath135 , with precision @xmath29 . the times
@xmath136 are defined as follows : putting @xmath137 , @xmath138 is given by the minimum time such that @xmath139 and so on . following the steps @xmath140 , the receiver can reconstruct , with a precision @xmath26 , the sequence @xmath141 , by simply iterating on a computer the evolution law ( [ eq : mappa ] ) between @xmath142 and @xmath143 , @xmath132 and @xmath144 , and so on .
the amount of bits necessary to implement the above transmission ( 1 - 4 ) can be easily computed . for simplicity of notation
we introduce the quantities @xmath145 which can be regarded as a sort of _ effective _ lyapunov exponents @xcite .
the le @xmath31 can be written in terms of @xmath146 as follows @xmath147 where @xmath148 is the average time after which we have to transmit the new initial condition .
note that to obtain @xmath31 from the @xmath149 s requires the average ( [ eq : liapt ] ) , because the transmission time , @xmath150 , is not constant . if @xmath58 is large enough
the number of transmissions , @xmath37 , is @xmath151 .
therefore , noting that in each transmission , a reduction of the error from @xmath26 to @xmath29 requires the employ of @xmath152 bits , the total amount of bits used in the transmission is @xmath153 in other words the number of bits for unit time is proportional to @xmath31 . in more than one dimension , we have simply to replace @xmath31 with @xmath83 in ( [ eq : bits ] ) , because the above transmission procedure has to be repeated for each of the expanding directions .
lyapunov exponents and ks - entropy are properly defined only in specific asymptotic limits : very long times and arbitrary accuracy . however , predictability problem in realistic situations entails considering finite time intervals and limited accuracy .
the first obvious way for quantifying the predictability of a physical system is in terms of the _ predictability time _ @xmath28 , i.e. the time interval on which one can typically forecast the system .
a simple argument suggests @xmath154 however , the above relation is too naive to be of practical relevance , in any realistic system . indeed , it does not take into account some basic features of dynamical systems . the lyapunov exponent is a global quantity , because it measures the average rate of divergence of nearby trajectories .
in general there exist finite - time fluctuations and their probability distribution functions ( pdf ) is important for the characterization of predictability .
generalized lyapunov exponents _ have been introduced with the purpose to take into account such fluctuations @xcite . moreover ,
the lyapunov exponent is defined for the linearized dynamics , i.e. , by computing the rate of separation of two infinitesimally close trajectories . on the other hand , in measuring the predictability time ( [ eq:2.1 - 1 ] )
one is interested in a finite tolerance @xmath26 , because the initial error @xmath29 is finite .
a recent generalization of the lyapunov exponent to _ finite size _
errors extends the study of the perturbation growth to the nonlinear regime , i.e. both @xmath29 and @xmath26 are not infinitesimal @xcite .
we discuss now an example where the lyapunov exponent is of little relevance for characterizing the predictability .
this problem can be illustrated by considering the following coupled map model : @xmath155 where @xmath156 , @xmath157 , @xmath158 is a rotation matrix of arbitrary angle @xmath159 , @xmath160 is a vector function and @xmath161 is a chaotic map . for simplicity
we consider a linear coupling @xmath162 and the logistic map @xmath163 . for @xmath164
we have two independent systems : a regular and a chaotic one .
thus the lyapunov exponent of the @xmath126 subsystem is @xmath165 , i.e. , it is completely predictable . on the contrary
, the @xmath166 subsystem is chaotic with @xmath167 .
the switching on of a small coupling ( @xmath168 ) yields a single three - dimensional chaotic system with a positive global lyapunov exponent @xmath169 a direct application of ( [ eq:2.1 - 1 ] ) would give @xmath170 but this result is clearly unacceptable : the predictability time for @xmath126 seems to be independent of the value of the coupling @xmath171 .
this is not due to an artifact of the chosen example , indeed , the same argument applies to many physical situations @xcite .
a well known example is the gravitational three body problem , with one body ( asteroid ) much smaller than the other two ( planets ) .
when the gravitational feedback of the asteroid on the two planets is neglected ( restricted problem ) , one has a chaotic asteroid in the regular field of the planets . as soon as the feedback is taken into account ( i.e. @xmath168 in the example ) one has a non - separable three body system with a positive le .
of course , intuition correctly suggests that , in the limit of small asteroid mass ( @xmath172 ) , a forecast of the planet motion should be possible even for very long times .
the apparent paradox arises from the misuse of formula ( [ eq:2.1 - 1 ] ) , strictly valid for tangent vectors , to the case of non infinitesimal regimes .
as soon as the errors become large , the full nonlinear evolution of the three body system has to be taken into account .
this situation is clearly illustrated by the model ( [ eq:2.3 - 1 ] ) in figure [ fig:2.3 - 1 ] .
the evolution of @xmath173 is given by @xmath174 where , with our choice , @xmath175 . at the beginning ,
both @xmath176 and @xmath177 grow exponentially . however , the available phase space for @xmath166 is finite and the uncertainty reaches the saturation value @xmath178 in a time @xmath179 . at larger times
the two realizations of the @xmath166 variable are completely uncorrelated and their difference @xmath177 in ( [ eq:2.3 - 4 ] ) acts as a noisy term . as a consequence ,
the growth of the uncertainty on @xmath126 becomes diffusive with a diffusion coefficient proportional to @xmath180 @xcite @xmath181 so that : @xmath182 this example shows that , even in simple systems , the lyapunov exponent can be of little relevance for the characterization of the predictability . in more complex systems , in which different scales are present , one is typically interested in forecasting the large scale motion , while the le is related to the small scale dynamics .
a familiar example of that is weather forecast : despite the le of the atmosphere is indeed rather large , due to the small scale convective motion , large - scale weather predictions are possible for about @xmath183 days @xcite .
it is thus natural to seek for a generalization of the le to finite perturbations from which one can obtain a more realistic estimation for the predictability time .
it is worth underlining the important fact that finite errors are not confined in the tangent space but are governed by the complete nonlinear dynamics . in this sense
the extension of the le to finite errors will give more information on the system .
aiming to generalize the le to non infinitesimal perturbations let us now define the finite size lyapunov exponent ( fsle ) @xcite .
consider a reference @xmath4 and a perturbed trajectory @xmath184 , such that @xmath185 .
one integrates the two trajectories and computes the time @xmath186 necessary for the separation @xmath187 to grow from @xmath188 to @xmath189 . at time
@xmath190 the distance between the trajectories is rescaled to @xmath188 and the procedure is repeated in order to compute @xmath191 .
the threshold ratio @xmath192 must be @xmath193 , but not too large in order to avoid contributions from different scales in @xmath194 .
a typical choice is @xmath195 ( for which @xmath194 is properly a `` doubling '' time ) or @xmath196 .
in the same spirit of the discussion leading to eq.s ( [ eq : mars1 ] ) and ( [ eq : liapt ] ) , we may introduce an effective finite size growth rate : @xmath197 after having performed @xmath198 error - doubling experiments , we can define the fsle as @xmath199 where @xmath200 is @xmath201 see @xcite for details . in the infinitesimal limit , the fsle reduces to the standard lyapunov exponent @xmath202 in practice this limit means that @xmath203 displays a constant plateau at @xmath204 for sufficiently small @xmath188 ( fig . [ fig:2.3 - 2 ] ) . for finite value of @xmath188 the behavior of @xmath203
depends on the details of the non linear dynamics .
for example , in the model ( [ eq:2.3 - 1 ] ) the diffusive behavior ( [ eq:2.3 - 5 ] ) , by simple dimensional arguments , corresponds to @xmath205 .
@xmath203 as a function of @xmath188 for the coupled map ( [ eq:2.3 - 1 ] ) with @xmath206 .
the perturbation has been initialized as in fig .
[ fig:2.3 - 1 ] . for @xmath207 , @xmath208 ( horizontal line ) .
the dashed line shows the behavior @xmath205 . ]
since the fsle measures the rate of divergence of trajectories at finite errors , one might wonder whether it is just another way to look at the average response @xmath209 as a function of time .
the answer is negative , because taking the average at fixed time is not the same as computing the average doubling time at _ fixed scale _ , as in ( [ eq:2.3 - 10 ] ) .
this is particularly clear in the case of strongly intermittent system , in which @xmath210 can be very different in each realization . in
the presence of intermittency , averaging over different realizations at fixed times can produce a spurious regime due to the superposition of exponential and diffusive contributions by different samples at the same time @xcite .
the fsle method can be easily applied to data analysis @xcite . for other approaches addressing the problem of non - infinitesimal perturbations
see @xcite . for most systems ,
the computation of kolmogorov - sinai entropy ( [ eq : ks ] ) is practically impossible , because it involves the limit on arbitrary fine resolution and infinite times .
however , in the same philosophy of the fsle , by relaxing the requirement of arbitrary accuracy , one can introduce the @xmath18-entropy which measures the amount of information for reproducing a trajectory with finite accuracy @xmath18 in phase - space . roughly speaking
the @xmath18-entropy can be considered the counterpart , in information theory , of the fsle .
such a quantity was originally introduced by shannon @xcite , and by kolmogorov @xcite . recently
gaspard and wang @xcite made use of this concept to characterize a large variety of processes .
we start with a continuous - time variable @xmath211 , which represents the state of a @xmath17-dimensional system , we discretize the time by introducing an interval @xmath82 and we consider the new variable @xmath212 of course @xmath213 and it corresponds to the trajectory which lasts for a time @xmath214 . in data analysis ,
the space where the state of the system lives is unknown and usually only a scalar variable @xmath215 can be measured .
then , one considers vectors @xmath216 , that live in @xmath217 and allow a reconstruction of the original phase space , known as delay embedding in the literature @xcite , and it is a special case of ( [ eq:2 - 1 ] ) .
introduce now a partition of the phase space @xmath218 , using cells of edge @xmath18 in each of the @xmath17 directions .
since the region where a bounded motion evolves contains a finite number of cells , each @xmath219 can be coded into a word of length @xmath32 , out of a finite alphabet : @xmath220 where @xmath221 labels the cell in @xmath218 containing @xmath222 . from the time evolution one
obtains , under the hypothesis of ergodicity , the probabilities @xmath223 of the admissible words @xmath224 .
we can now introduce the @xmath225-entropy per unit time , @xmath226 @xcite : @xmath227 where @xmath228 is the block entropy of blocks ( words ) with length @xmath32 : @xmath229 for the sake of simplicity , we ignored the dependence on details of the partition . to make @xmath230 partition - independent one has to consider a generic partition of the phase space @xmath231 and to evaluate the shannon entropy on this partition : @xmath232 .
the @xmath171-entropy is thus defined as the infimum over all partitions for which the diameter of each cell is less than @xmath171 @xcite : @xmath233 note that the time dependence in ( [ def : eps ] ) is trivial for deterministic systems , and that in the limit @xmath234 one recovers the kolmogorov - sinai entropy @xmath235
in the previous sections , we discussed the characterization of dynamical behaviors when the evolution laws are known either exactly or with some degree of uncertainty . in experimental investigations , however , only time records of some observable are available , while the equations of motion for the observable are generally unknown . the predictability problem of this latter case , at least from a conceptual point of view ,
can be treated as if the evolution laws were known .
indeed , in principle , the embedding technique allows for a reconstruction of the phase space @xcite .
nevertheless there are rather severe limitations for high dimensional systems @xcite and even in low dimensional ones non trivial features appear in the presence of noise @xcite . in this section we show that an entropic analysis at different resolution scales provides a pragmatic classification of a signal and gives suggestions for modeling of systems . in particular we illustrate , using some examples , how quantities such as the @xmath18-entropy or the fsle can display a subtle transition from the large to the small scales .
a negative consequence of this is the difficulty in distinguishing , only from data analysis , a genuine deterministic chaotic system from one with intrinsic randomness @xcite . on the other hand ,
the way the @xmath18-entropy or fsle depends on the ( resolution ) scale , allows for a classification of the stochastic or chaotic character of a signal , and this gives some freedom in modeling the system .
the `` true character '' of the number sequence @xmath236 obtained by a ( pseudo ) random number generator ( prng ) on a computer is an issue of paramount importance in computer simulations and modeling .
one would like to have a sequence with a random character as much as possible , but is forced to use deterministic algorithms to generate @xmath236 .
this subsection is mainly based on the paper @xcite .
a simple and popular prng is the multiplicative congruent one : @xmath237 with an integer multiplier @xmath238 and modulus @xmath239 .
the @xmath240 are integer numbers from which one hopes to generate sequence of random variables @xmath241 , which are uncorrelated and uniformly distributed in the unit interval .
a first problem arises from the periodic nature of the rule ( [ prng ] ) as a consequence of its discrete nature .
note that the rule ( [ prng ] ) can be interpreted also as a deterministic dynamical system , i.e. @xmath242 which has a uniform invariant measure and a ks entropy @xmath243 .
when imposing the integer arithmetics of eq .
( [ prng ] ) onto this system , we are , in the language of dynamical systems , considering an unstable periodic orbit of eq .
( [ xdyn ] ) , with the particular constraint that , to achieve the period @xmath244 ( i.e. all integers @xmath245 should belong to the orbit of eq .
( [ prng ] ) ) , it has to contain all values @xmath246 , with @xmath247 . since the natural invariant measure of eq .
( [ xdyn ] ) is uniform , such an orbit represents the measure of a chaotic solution in an optimal way .
every sequence of a prng is characterized by two quantities : its period @xmath248 and its positive lyapunov exponent @xmath31 , which is identical to the entropy of a chaotic orbit of the equivalent dynamical system .
of course a good random number generator must have a very large period , and as large as possible entropy .
it is natural to ask how this apparent randomness can be reconciled with the facts that ( a ) the prng is a deterministic dynamical systems ( b ) it is a discrete state system .
if the period is long enough , on shorter times only point ( a ) matters and it can be discussed in terms of the behavior of the @xmath18-entropy , @xmath249 . at high resolutions ( @xmath250 )
, it seems rather reasonable to think that the true deterministic chaotic nature of the congruent rule shows up , and , therefore , @xmath251 . on the other hand , for @xmath252 , one expects to observe the `` apparent random '' behavior of the system , i.e. @xmath253 , see fig [ fig : rng_entropy ] .
the @xmath18-entropies , @xmath254 , at varying the embedding dimension @xmath32 for the multiplicative congruential random number generator eq . [ prng ] for different choices of @xmath238 and @xmath239 .
] we discuss an example of high - dimensional system with a non - trivial behavior at varying the resolution scales , namely the emergence of nontrivial collective behavior .
let us consider a globally coupled map ( gcm ) defined as follows @xmath255 where @xmath37 is the total number of elements , and @xmath256 is a chaotic map on the interval @xmath257 $ ] , depending on the control parameter @xmath258 .
the evolution of a macroscopic variable , e.g. , the center of mass @xmath259 upon varying @xmath171 and @xmath258 in eq .
( [ eq:3.38 ] ) , displays different behaviors @xcite : \(a ) _ standard chaos _
: @xmath260 obeys a gaussian statistics with a standard deviation @xmath261 ; \(b ) _ macroscopic periodicity _ : @xmath260 is a superposition of a periodic function and small fluctuations @xmath262 ; \(c ) _ macroscopic chaos _ : @xmath260 exhibits an irregular motion , as seen by plotting @xmath260 vs. @xmath263 .
the plot sketches a structured function ( with thickness @xmath264 ) , and suggests a chaotic motion for @xmath260 . in the case of _
macroscopic chaos _
, the center of mass is expected to evolve with typical times longer than the characteristic time @xmath265 of the full dynamics ( microscopic dynamics ) ; @xmath204 being the lyapunov exponent of the gcm . indeed ,
conceptually , macroscopic chaos for gcm can be thought of as the analogous of the hydro - dynamical chaos for molecular motion . in spite of a huge microscopic lyapunov exponent ( @xmath266
, @xmath267 is the collision time ) , one can have rather different behaviors at a hydro - dynamical ( coarse grained ) level : regular motion ( @xmath268 ) or chaotic motion ( @xmath269 ) . in principle , if the hydrodynamic equations were known , a characterization of the macroscopic behavior would be possible by means of standard dynamical system techniques .
however , in generic cml there are no general systematic methods to build up the macroscopic equations , apart from particular cases @xcite .
we recall that for chaotic systems , in the limit of infinitesimal perturbations @xmath207 , one has @xmath270 , i.e. @xmath203 displays a plateau at the value @xmath271 for sufficiently small @xmath188 . however , for non infinitesimal @xmath188 , one can expect that the @xmath188-dependence of @xmath203 may give information on the characteristic time - scales governing the system , and , hence , it could be able to characterize the macroscopic motion .
in particular , at large scales ( @xmath272 ) , the fast microscopic components saturate and @xmath273 , where @xmath274 can be fairly called the `` macroscopic '' lyapunov exponent .
the fsle has been determined by looking at the evolution of @xmath275 , which has been initialized at the value @xmath276 by shifting all the elements of the unperturbed system by the quantity @xmath277 ( i.e. @xmath278 ) , for each realization .
the computation has been performed by choosing the tent map as local map , but similar results can be obtained for other maps @xcite .
the main result can be summarized as follows : * at small @xmath279 , where @xmath37 is the number of elements , the `` microscopic '' lyapunov exponent is recovered , i.e. @xmath280 * at large @xmath281 , another plateau @xmath282 appears , which can be much smaller than the microscopic one .
the emerging scenario is that , at a coarse - grained level , i.e. @xmath272 , the system can be described by an `` effective '' hydro - dynamical equation ( which in some cases can be low - dimensional ) , while the `` true '' high - dimensional character appears only at very high resolution , i.e. @xmath283 consider the following map which generates a diffusive behavior on the large scales @xcite : @xmath284 where @xmath285 indicates the integer part of @xmath286 and @xmath287 is given by : @xmath288\ , .
\label{eq : mappaf}\ ] ] the largest lyapunov exponent @xmath31 can be obtained immediately : @xmath289 , with @xmath290 .
one expects the following scenario for @xmath249 : @xmath291 @xmath292 where @xmath293 is the diffusion coefficient , @xmath294 .
the map @xmath295 ( [ eq : mappaf ] ) for @xmath296 is shown with superimposed the approximating ( regular ) map @xmath297 ( [ eq:3 - 5 ] ) obtained by using @xmath298 intervals of slope @xmath299 . ]
consider now a stochastic system , namely a noisy map @xmath300 where @xmath301 , as shown in fig . [ map ] , is a piece wise linear map which approximates the map @xmath287 , and @xmath302 is a stochastic process uniformly distributed in the interval @xmath303 , and no correlation in time . when @xmath304 , as is the case we consider , the map ( [ eq:3 - 5 ] ) , in the absence of noise , gives a non - chaotic time evolution .
lyapunov exponent @xmath305 versus @xmath18 obtained for the map @xmath287 ( [ eq : mappaf ] ) with @xmath296 ( @xmath306 ) and for the noisy ( regular ) map ( [ eq:3 - 5 ] ) ( @xmath307 ) with @xmath308 intervals of slope @xmath309 and @xmath310 .
straight lines indicate the lyapunov exponent @xmath311 and the diffusive behavior @xmath312 .
now we compare the finite size lyapunov exponent for the chaotic map ( [ eq:3 - 1 ] ) and for the noisy one ( [ eq:3 - 5 ] ) . in the latter
the fsle has been computed using two different realizations of the noise . in fig .
[ fslediff ] we show @xmath305 versus @xmath18 for the two cases .
the two curves are practically indistinguishable in the region @xmath313 .
the differences appear only at very small scales @xmath314 where one has a @xmath305 which grows with @xmath18 for the noisy case , remaining at the same value for the chaotic deterministic case .
both the fsle and the @xmath315-entropy analysis show that we can distinguish three different regimes observing the dynamics of ( [ eq:3 - 5 ] ) on different length scales . on the large length scales
@xmath316 we observe diffusive behavior in both models .
on length scales @xmath317 both models show chaotic deterministic behavior , because the entropy and the fsle are independent of @xmath18 and larger than zero .
finally on the smallest length scales @xmath318 we see stochastic behavior for the system ( [ eq:3 - 5 ] ) , _ i.e. _ @xmath319 , while the system ( [ eq:3 - 1 ] ) still shows chaotic behavior .
the above examples show that the distinction between chaos and noise can be a highly non trivial task , which makes sense only in very peculiar cases , e.g. , very low dimensional systems .
nevertheless , even in this case , the entropic analysis can be unable to recognize the `` true '' character of the system due to the lack of resolution .
again , the comparison between the diffusive map ( [ eq:3 - 1 ] ) and the noisy map ( [ eq:3 - 5 ] ) is an example of these difficulties . for @xmath320 both the system ( [ eq:3 - 1 ] ) and ( [ eq:3 - 5 ] ) , in spite of their `` true '' character , will be classified as chaotic , while for @xmath321 both can be considered as stochastic . in high - dimensional chaotic systems , with @xmath37 degrees of freedom
, one has typically @xmath322 for @xmath323 ( where @xmath324 as @xmath325 ) while for @xmath326 , @xmath249 decreases , often with a power law @xcite .
since also in some stochastic processes the @xmath18-entropy obeys a power law , this can be a source of confusion .
these kind of problems are not abstract ones , as a recent debate on `` microscopic chaos '' demonstrates @xcite .
the detection of microscopic chaos by data analysis has been recently addressed in a work of gaspard et al .
these authors , from an entropic analysis of an ingenious experiment on the position of a brownian particle in a liquid , claim to give an empirical evidence for microscopic chaos .
in other words , they state that the diffusive behavior observed for a brownian particle is the consequence of chaos at a molecular level .
their work can be briefly summarized as follows : from a long ( @xmath327 data ) record of the position of a brownian particle they compute the @xmath18-entropy with the cohen - procaccia method @xcite from which they obtain : @xmath328 where @xmath293 is the diffusion coefficient .
then , _ assuming _ that the system is deterministic , and making use of the inequality @xmath329 , they conclude that the system is chaotic
. however , their result does not give a direct evidence that the system is deterministic and chaotic .
indeed , the power law ( [ eq : gasp ] ) can be produced with different mechanisms : 1 . a genuine chaotic system with diffusive behavior , as the map ( [ eq : mappaf ] ) ; 2 . a non chaotic system with some noise , as the map ( [
eq:3 - 5 ] ) , or a genuine brownian system ; 3 .
a deterministic linear non chaotic system with many degrees of freedom ( see for instance @xcite ) ; 4 . a `` complicated '' non chaotic system as the ehrenfest wind - tree model where a particle diffuses in a plane due to collisions with randomly placed , fixed oriented square scatters , as discussed by cohen et al .
@xcite in their comment to ref .
@xcite .
it seems to us that the weak points of the analysis in ref .
@xcite are : \a ) the explicit assumption that the system is deterministic ; \b ) the limited number of data points and therefore limitations in both resolution and block length .
the point ( a ) is crucial , without this assumption ( even with an enormous data set ) it is not possible to distinguish between 1 ) and 2 ) .
one has to say that in the cases 3 ) and 4 ) at least in principle it is possible to understand that the systems are `` trivial '' ( i.e. not chaotic ) but for this one has to use a huge number of data .
for example cohen et al .
@xcite estimated that in order to distinguish between 1 ) and 4 ) using realistic parameters of a typical liquid , the number of data points required has to be at least @xmath330 . concluding , we have the apparently paradoxical result that `` complexity '' helps in the construction of models .
basically , in the case in which one has a variety of behaviors at varying the scale resolution , there is a certain freedom on the choice of the model to adopt .
for some systems the behavior at large scales can be realized both with chaotic deterministic models or suitable stochastic processes . from a pragmatic point of view ,
the fact that in certain stochastic processes @xmath331 can be indeed extremely useful for modeling such high - dimensional systems .
perhaps , the most relevant case in which one can use this freedom in modeling is the fully developed turbulence whose non infinitesimal ( the so - called inertial range ) properties can be successfully mimicked in terms of multi - affine stochastic process ( see ref .
the guideline of this paper has been _ the interpretation of different aspects of the predictability of a system as a way to characterize its complexity_. we have discussed the relation between chaoticity , the kolmogorov - sinai entropy and algorithmic complexity . as clearly exposed in the seminal works of alekseev and yakobson @xcite and ford @xcite , the time sequences generated by a system with sensitive dependence on initial conditions have non - zero algorithmic complexity .
a relation exists between the maximal compression of a sequence and its ks - entropy .
therefore , one can give a definition of complexity , without referring to a specific description , as an intrinsic property of the system .
the study of these different aspects of predictability constitutes a useful method for a quantitative characterization of `` complexity '' , suggesting the following equivalences : @xmath332 the above point of view , based on dynamical systems and information theory , quantifies the complexity of a sequence considering each symbol relevant but it does not capture the structural level .
let us clarify this point with the following example .
a binary sequence obtained with a coin tossing is , from the point of view adopted in this review , complex since it can not be compressed ( i.e. it is unpredictable ) . on the other hand
such a sequence is somehow trivial , i.e. with low `` organizational '' complexity .
it would be important to introduce a quantitative measure of this intuitive idea .
the progresses of the research on this intriguing and difficult issue are still rather slow .
we just mention some of the most promising proposals as the logical depth and the sophistication @xcite .
f. takens , `` detecting strange attractors in turbulence '' in _ dynamical systems and turbulence ( warwick 1980 ) _ , vol .
898 of _ lecture notes in mathematics _ ,
rand and l .- s .
young ( eds . ) , pg . 366 , springer - verlag , berlin ( 1980 ) . | some aspects of the predictability problem in dynamical systems are reviewed .
the deep relation among lyapunov exponents , kolmogorov - sinai entropy , shannon entropy and algorithmic complexity is discussed . in particular , we emphasize how a characterization of the unpredictability of a system gives a measure of its complexity .
a special attention is devoted to finite - resolution effects on predictability , which can be accounted with suitable generalization of the standard indicators .
the problems involved in systems with intrinsic randomness is discussed , with emphasis on the important problems of distinguishing chaos from noise and of modeling the system .
_ all the simple systems are simple in the same way , each complex system has its own complexity _ ( freely inspired by _ anna karenina _ by lev n. tolstoy ) |
the study of rare kaon decays @xcite has played an important role in the establishment of the standard model ( sm ) in particle physics .
it has also been crucial for understanding the phenomenon of cp violation @xcite .
the rare decay @xcite is a direct cp violation process caused by a flavor - changing neutral current ( fcnc ) with transition from a strange to down quark .
the unique characteristic of the decay is that the branching ratio can be calculated with very small theoretical uncertainties . in the sm based on the cabibbo - kobayashi - maskawa ( ckm ) matrix @xcite for quark flavor mixing
, the branching ratio is predicted to be @xmath1 @xcite .
the uncertainty of the prediction is dominated by the allowed range of the imaginary part of a ckm matrix element @xmath2 , which is determined by other measurements , and the intrinsic theoretical uncertainty is only 1 - 2% @xcite . the determination of ckm matrix elements has been greatly improved in the past ten years by measuring various b - decay properties @xcite ; all measurements were consistent with each other within the standard model parameterization . by measuring the decay precisely , we can check the consistency against the currently predicted value in the sm .
a deviation indicates new physics beyond the sm because of the very small uncertainty in deriving the imaginary part of @xmath2 from the branching ratio of the decay .
the decay is one of the processes expected to have a significant impact on new physics searches because it is an fcnc process that can proceed through loop diagrams , including the interactions at short distance and large mass scales .
grossman and nir @xcite pointed out the importance of studying new physics by using both the and decays , where the decay is another rare kaon decay with small theoretical uncertainties @xcite .
various new physics models have been developed and used to predict the and branching ratios @xcite .
the best upper limit obtained in previous experiments was br ( ) @xmath3 5.9 @xmath4 10@xmath5 ( 90% cl ) @xcite .
the ultimate goal of our experimental study is to determine the branching ratio with an accuracy less than 10% of the value predicted in the sm .
the goal can be achieved by performing a series of experiments with improved and refined detection methods .
the e391a experiment , which was carried out at the kek 12-gev proton synchrotron ( kek - ps ) , is the first step of this approach .
the main objectives of e391a were not only to investigate the decay with a dedicated detector , but also to test and confirm our basic experimental methods for at the highest possible sensitivity .
data collection of the e391a experiment started in february 2004 and continued until november 2005 , which was one month before the shutdown of the kek - ps .
the total running time was about twelve months and it was divided into three periods ( run-1 , run-2 , and run-3 ) .
early results from the first and second periods were reported in ref .
@xcite and ref .
@xcite , respectively . in this article
, we report the final results on the decay as obtained from e391a .
the key signature of the decay is detection of exactly two photons and nothing else .
the decays dominantly into two photons ( @xmath6 ) , and the neutrinos are undetectable . because detection of the incident is difficult , measurements can only be made of the energy and position of the two photons in a calorimeter located downstream of the decay region , without having direct information of the incident particle .
kinematic constraints are weak for definitive identification of the decay . instead
, the decay must be isolated by eliminating all possible backgrounds .
because the signal mode is identified as the final state of two photons and nothing else , processes that make two or more photons can cause background events .
the decays such as and become backgrounds when extra photons or charged particles escape detection .
the decay can be a background source because it has only two photons in the final state , although it is well suppressed by kinematical constraints .
another background process is hadronic interactions of beam neutrons with the residual gas in the beamline or in detectors near the beam .
any or @xmath7 produced in these interactions decay into two photons and produce backgrounds .
hyperons produced at the target can cause backgrounds through processes such as @xmath8 ; these hyperon events are strongly suppressed because most of them decay in a 10 m long neutral beamline . the most important tool for reducing the background is a hermetic detector system to detect and veto extra particles .
because all of the other decays , except for , are accompanied with at least two additional photons or charged particles , the detector should be highly sensitive to photons and charged particles .
the signature of can be provided by using a small - diameter beam ( called a pencil beam " ) and measuring precisely the energy and position of the two photons .
although flux is reduced with a pencil beam , it has several advantages .
first , the beam hole at the center of the calorimeter , which compromises hermeticity , can be minimized .
secondly , the decay vertex position ( ) can be assumed to be on the beam axis .
the vertex position , which is same as the decay position due to the short lifetime of , is obtained from the kinematics of the two photons from its decay ( see sec .
the transverse momentum of the ( ) with respect to the beam axis is also obtained .
the signal region for decay can be defined with and .
requiring a sufficiently large missing transverse momentum eliminates contamination from the decay and also reduces the contamination of other decays .
most of the s decay into multiple particles that have low momenta in the rest frame , and hence low in the laboratory frame .
the should be in the region away from beam counters and should be in the range from 120 to 240 mev/@xmath9 , where the maximum momentum of s in the -rest frame is 231 mev/@xmath9 for .
production of s through beam - gas interactions in the decay fiducial region are reduced by having a high vacuum .
hit - rates of the beam halo ( mostly neutrons ) with the surrounding detectors are minimized by a sharp collimation of the beam .
in addition , as discussed in sec .
iii.a and in the subsequent analysis of the data , mis - combinations of two photons from background processes and/or mis - measurements of energies and positions were found to be major sources of backgrounds in the e391a experiment .
the photons were often produced from interactions of the beam halo and can lead to incorrect determinations of .
some backgrounds from beam interactions can be reduced by detecting low - energy deposits from recoil particles emitted after the interactions .
more detailed descriptions of the detection method have been reported in the original proposal of e391a @xcite .
inefficiencies of photon and charged particle detection provide backgrounds .
the decay is the most serious background source arising from an inefficiency because it has only two extra photons in the final state .
the inefficiency of photon detection was investigated in a series of experiments @xcite , which showed that the inefficiency monotonically decreased with the incident photon energy .
it was also found that photon detection with a very low energy threshold is necessary to achieve a small overall inefficiency , even for high - energy photons .
if an extremely low energy threshold around 1 mev was set , the backgrounds from other decays were reduced to a negligible level within the sensitivity of the experiment .
the following sections describe the application of the basic methods in the e391a experiment .
the production target of the beam was a platinum rod with a length of 60 mm and a diameter of 8 mm .
the profile of the primary beam was @xmath10 = 3.3 mm and @xmath10 = 1.1 mm along the @xmath11 ( horizontal ) and @xmath12 ( vertical ) axes .
the neutral beam was extracted at an angle of 4@xmath13 with respect to the primary proton beam .
the target rod and beamline elements were aligned along a straight line in the direction of the neutral beam .
the total length of the beamline was 10 m. the neutral beamline consisted of two dipole magnets ( d1 , d2 ) to sweep charged particles out of the beam , with field strengths of 4 t@xmath14 m and 3 t@xmath14 m , respectively , and six collimators ( c1 - c6 ) to collimate the beam , as shown in fig .
[ fig_beamline ] .
coordinate is at the center of the production target . ]
five collimators ( c1-c3 , c5 , and c6 ) were made of tungsten .
the c2 and c3 collimators in the upstream end were used to define the beam with the designed apertures , which were arranged to form a half cone angle of 2 mrad from the target center .
the last two collimators , c5 and c6 , were used to trim the beam halo .
the most upstream collimator c1 reduced the size of the beam immediately after the target without producing a large penumbra .
the total thickness of these five collimators was approximately 6 m. a thermal - neutron absorber , which was made of polyethylene terephthalate ( pet ) sheets containing gd@xmath15o@xmath16 40% in weight , was used for c4 .
the aperture of c4 was set to be larger than that of the other collimators .
movable absorbers , made of lead ( pb ) and beryllium ( be ) , were placed between c1 and c2 to reduce the number of photons and neutrons relative to the s .
the absorbers were 10-mm - diameter rods with the lengths of 5 cm and 30 cm for pb and be , respectively .
the downstream region , starting with a stainless steel window 100 @xmath17 m thick at the upstream end of c4 , was evacuated to approximately 1 pa . the primary beam on the target was monitored by a secondary - emission chamber ( sec ) placed upstream of the target , and a target monitor ( tm ) that was a counter telescope which viewed the target center at 90 degrees .
the primary beam position at the target was adjusted with steering magnets by monitoring the decay events . in the @xmath18 events ,
six photons were detected by a csi calorimeter which will be described in sec .
the center of energy was defined to be @xmath19 , where @xmath20 are the photon energies and @xmath21 are the photon hit positions at the csi calorimeter . because the six - photon events were mostly decays , the peak position of the center of energy should be on the axis of the beamline .
the position of the neutral beam was maintained to be within 0.2 mm from the center throughout the entire running period .
the beam size at the csi calorimeter , was also monitored by the distribution of the center of energy .
its diameter was @xmath22 mm , which was consistent with the beam divergence .
the peak momentum of the as determined from beamline simulations was 2 gev/@xmath9 at the exit of c6 , as shown in fig .
[ fig_spectrum ] .
the initial spread due to the beam divergence of 2 mrad was approximately 4 mev/@xmath9 .
the neutron - to - ratio was 60 and the halo - to - core ratio was approximately 10@xmath23 for both neutrons and photons with energies above 1 mev , as shown in fig .
[ fig_profile ] . by inserting a
be absorber , the neutron - to - ratio was reduced to 40 , with roughly a 45% loss in flux .
profiles of the beam are shown in fig .
[ fig_klprofile ] .
punch - through muons were emitted in the direction parallel to the beam axis .
their position distribution was almost flat and the flux density was larger than the cosmic ray flux by roughly one order of magnitude .
details of the beamline has been reported elsewhere @xcite .
the e391a detection system was located at the end of the beamline .
the detector subsystems were cylindrically arranged around the beam axis , and most of them were placed inside a large vacuum vessel , as shown in fig .
[ fig_apparatus ] . from here on , the origin of the coordinate system is defined to be at the upstream end of the e391a detector , as shown in fig .
[ fig_apparatus ] .
this position was approximately 12 m from the production target .
the energy and hit position of photons , such as the two photons that would come from the decay in the signal , were measured by using a calorimeter placed at the downstream end of the decay region .
as shown in fig .
[ fig_csi ] , the calorimeter was made of 496 undoped csi crystals each with a dimension of 7 @xmath4 7 @xmath4 30 @xmath24 ( main csi ) @xcite and assembled in a cylindrical shape with an outer diameter of 1.9 m. there was a 12 cm @xmath4 12 cm beam hole at the center of the calorimeter .
the beam hole was surrounded by a tungsten - scintillator collar counter ( cc03 ) and 24 csi modules with dimensions of 5 @xmath4 5 @xmath4 50 @xmath24 ( ktev csi ) @xcite .
the gaps at the periphery of the cylinder were filled with 56 csi modules with trapezoidal shape and 24 modules of lead - scintillator sandwich counters ( lead / scintillator sandwich ) .
these modules were tightly stacked such that the gaps between them were less than 0.1 mm .
although the cc03 and the sandwich counters were used only in the veto system , the other parts , which consisted of csi crystals , were used for photon measurements .
the average light yield of the main csi modules was 16 pe / mev , where pe / mev is the unit of the number of photo - electrons emitted from the photomultiplier tube ( pmt ) cathode for an energy deposit of 1 mev in the detector .
details of the csi calorimeter have been reported in ref .
@xcite .
a set of plastic scintillation counters ( named cv ) were placed in front of the csi calorimeter to identify events that included the emission of charged particles , such as @xmath25 decay @xcite .
a total of 32 sector - shaped modules ( outer cv ) were placed at a distance of 50 cm from the front face of the csi calorimeter .
the outer cv modules were arranged to have overlaps between adjacent modules .
the beam region from the outer cv to the csi was covered by the inner cv , which was a square pipe formed with four plastic scintillator plates .
the inner and outer cvs were closely connected with aluminum fixtures . in order to eliminate gaps between the outer and inner modules ,
the inner modules were extended to cover the edge of the outer modules .
the edges of the outer cv and the inner cv were located close to the beam , and neutrons in the beam halo frequently interacted with them . the decay region was surrounded with two large lead - scintillator sandwich counters : the main barrel ( mb ) and the front barrel ( fb ) , composed of 32 and 16 modules , respectively .
the modules were tightly assembled with small gaps @xcite , as shown in fig .
[ fig_barrels ] .
the lamination was parallel to the beam axis , and the lengths of the mb and fb modules were 5.5 m and 2.75 m , respectively .
we developed a new type of plastic scintillator made of a resin mechanically strengthened with a mixture of styrene and methacrylate @xcite .
the scintillator counters were fabricated by extrusion .
the thickness of the mb and fb modules were 13.5@xmath26 and 17.2@xmath26 , respectively .
the emitted light was transmitted through wavelength - shifting ( wls ) fibers , which were glued to each scintillator plate with a pitch of 10 mm .
the fibers had a double - cladded structure with y11 as a dopant .
the green light emitted from y11 was read by a newly developed photomultiplier tube with a high quantum efficiency for green light ( egp - pmt ) @xcite . for readout , each module was divided into an outer and an inner parts .
the mb module was viewed from both ends , and the fb module was viewed from the upstream end .
figure [ fig_mblight ] shows the light yields for 4 readout channels of one mb module , obtained from a cosmic - ray test .
the light yield and attenuation of the fb module were similar to those shown in fig .
[ fig_mblight ] for the mb module .
details of the barrel counters has been reported in ref .
@xcite .
a layer of plastic scintillator was placed inside of the mb modules to identify charged particles .
it consisted of 32 modules and was called the barrel charged veto ( bcv ) .
a module was made of a 10-mm - thick plastic scintillator , and the light signal was read from both ends by an egp - pmt through wls fibers glued to the scintillator with a 5-mm pitch .
multiple collar - shaped counters , cc02
cc07 , were placed along the beam axis .
counter cc02 at the entrance of the decay region was a lead - scintillator sandwich of the shashlik type , with wls fibers piercing the lead and scintillator layers .
counter cc03 at the end of the decay region was a tungsten - scintillator sandwich .
counters cc04 cc07 covered the solid - angle in the downstream direction . counters cc04 and
cc05 were lead - scintillator sandwiches with wls fibers glued on each scintillator plate at a pitch of 10 mm , which was the same as for the mb and fb .
counters cc06 and cc07 were made of sf5 lead - glass .
the direction of lamination was perpendicular to the beam for cc02 , cc04 , and cc05 , and was parallel to the beam for cc03 .
the signals from cc02 cc05 were read by egp - pmts . beginning with run-2 , another collar counter called cc00
was installed in front of the fb and outside the vacuum vessel to reduce the effects of halo neutrons .
it consisted of 11 layers of a 5-mm - thick plastic scintillator interleaved with 10 layers of 20-mm - thick tungsten .
however , cc00 did not significantly reduce the neutron halo because it was not placed as close to the beam as the other collar counters .
a beam - plug counter , back - anti ( ba ) , was placed at the end of detector system along the beam axis . in the first two data - taking periods , run-1 and run-2 ,
the ba consisted of six superlayers , where each superlayer had six plastic - scintillator layers interleaved with lead sheets , and a single layer of quartz . in run-3 ,
the lead and plastic layers were replaced with pwo crystals as shown in fig [ fig_ba ] , with the intention to separate electromagnetic showers from the neutron hits . a thin layer of plastic scintillator , the beam hole charged veto ( bhcv ) , was placed in front of the ba .
it was effective in removing the @xmath27 decay .
the thickness of the bhcv was 1 mm for run-1 , and 3 mm for run-2 and run-3 .
+ the upstream counters , cc02
cc04 , were placed inside the vacuum vessel , and the downstream cc05 ba were placed outside the vessel .
a vacuum duct , which was directly connected to the vacuum vessel , penetrated cc05 cc07 and the vacuum region was extended up to the front face of bhcv . the size and basic parameters of cc02
cc07 , bhcv , and ba are summarized in table [ tab : table1 ] . [
cols="<,^,^,^,<,^,^,^",options="header " , ] the acceptance for the decay was estimated from monte - carlo simulations .
the raw acceptance was calculated by dividing the number of remaining events after imposing all the analysis cuts by the number of decays generated in the simulation .
it was 1.40% for the run-2 and 1.39% for the run-3 data , respectively . in the raw acceptance , the losses caused by geometrical acceptance , veto cuts , kinematic selections , and selection on fiducial region were included .
the geometrical acceptance to detect two photons in the csi was approximately 20% .
the loss by veto cuts , or self - vetoing , was about 50% , which was dominated by the csi and mb vetoes .
the loss in the csi veto was caused by hits accompanied by the genuine photon cluster , as shown in fig .
[ fig_csiveto ] , and the loss in the mb was caused by photons escaping from the front face of the csi and hitting the mb .
the acceptance of the kinematic and signal region selections was approximately 15% .
these values were estimated through mc simulations of the mode .
the evaluation of the acceptance loss was supported by the fact that the acceptance losses in the normalization modes ( , , and ) were reproduced by the simulation .
the acceptance loss due to accidental activities in the detector was estimated from real data taken with the tm trigger .
the accidental loss was estimated to be 20.6% for the run-3 data , in which the losses in mb ( 7.4% ) and ba ( 6.4% ) were major contributions .
for run-2 , the accidental loss was estimated to be 17.4% ; the difference between run-2 and run-3 was due to the difference in the ba counters used in the data taking .
the acceptance loss caused by the selections on the timing dispersion of each photon cluster and on the timing difference between two photons was estimated separately by using real data and was obtained to be 8.9% .
thus , the total acceptance for the was @xmath28% for run-2 and @xmath29% for run-3 case , where the errors are dominated by the systematic uncertainties that are discussed later .
figure [ fig_ptz_signal ] shows the distribution of the mc events in the scatter plot of -after imposing all of the other cuts .
position for the monte carlo events after imposing all of the analysis cuts .
the box indicates the signal region for . ] by using the number of decays and the total acceptance , the single event sensitivity for was @xmath30 for run-2 , @xmath31 for run-3 , and @xmath32 in total . after finalizing all of the event selection cuts ,
the candidate events inside the signal region were examined .
no events were observed in the signal region , as shown in fig .
[ fig_final ] .
an upper limit for the branching ratio was set to be @xmath33 at the 90% confidence level , based on poisson statistics . the result improves the limit previously published @xcite by a factor of 2.6 .
position for the events with all of the selection cuts imposed .
the box indicates the signal region for . ]
although the systematic uncertainties were not taken into account in setting the current upper limit on the branching ratio , we will describe our treatment for them to provide a thorough understanding of the experiment .
in particular , systematic uncertainties of the single event sensitivity and background estimates due to halo neutrons are discussed in order .
the systematic uncertainty of the single event sensitivity was evaluated by summing the uncertainties of the number of decays and the acceptance of the decay . because the calculation of the former also includes the acceptance of the normalization modes , the acceptance of both the normalization mode and the mode are relevant to the acceptance evaluation by the monte carlo simulations . to estimate the uncertainties in the acceptance calculation
, we utilized the fractional difference between data and the simulation in each selection criterion , defined by the equation @xmath34 where @xmath35 and @xmath36 denote the acceptance values of the @xmath37-th cut , calculated as the ratio of numbers of events with and without the cut , for the data and mc simulations , respectively . in @xmath38 ,
the acceptance was calculated with all the other cuts imposed .
the systematic uncertainty of the acceptance was evaluated by summing all the fractional differences in quadrature , weighted by the effectiveness of each cut , as @xmath39 for the three decay modes used in the normalization , , , and , the calculated uncertainties were 5.2% , 5.7% , and 3.6% ( in run-3 ) , respectively .
the acceptances of the csi veto cut had the largest uncertainties in all decay modes .
the number of decays was obtained by using the mode , and thus its uncertainty is quoted as the same value for the mode . for the acceptance of the mode ,
the same systematic uncertainties as the mode were adopted because there were no signal candidates in the data to be compared with the mc simulations .
the systematic error of the single event sensitivity was evaluated to be a quadratic sum of the uncertainties of the number of decays and the acceptance of the decay .
it was 10.3% in run-2 and 8.2% in run-3 , respectively .
the systematic errors of halo neutron backgrounds were also estimated by utilizing fractional differences between data and the simulations .
there were two kinds of cuts in our analysis , veto cuts and kinematic selections .
for the former , the method was unable to be used directly because veto cuts had been applied in the early stages of the mc simulations to save the computing time .
thus , for the systematic uncertainty due to veto cuts , the same value obtained in the analysis was assigned .
for the kinematic selections , the same method as described in the previous section was used .
the acceptance of each cut was calculated with all veto cuts imposed except kinematic selections . in total , the systematic uncertainties due to fractional differences were calculated to be 31% , 32% , and 44% for cc02- , cv- , and cv-@xmath7 backgrounds , respectively .
in addition , the uncertainties in the normalizations of the mc simulations were taken into account .
because the normalization was determined by using the number of events in the cc02 region ( region-(1 ) ) , there can be ambiguity in estimating the cv - related backgrounds . as shown in table [ table_outside ] , there was a 24% difference between data and mc simulation in the downstream region ( region-(2 ) ) .
even though they were statistically consistent , we assigned the difference as an additional systematic uncertainty of cv- . for the cv-@xmath7 case , further ambiguity due to the reproducibility of @xmath7 production should be considered .
it was estimated to be 24% from the difference between data and mc simulations in the numbers of @xmath7 events in the al plate run , as shown in fig .
[ fig_alrun ] .
the total systematic uncertainties were calculated by summing up the contributions quadratically , to be 31% , 40% , and 55% for cc02- , cv- , and cv-@xmath7 backgrounds , respectively .
the e391a experiment at the kek - ps was the first dedicated experiment for the decay . combining the periods of run-2 and run-3 ,
the single event sensitivity reached @xmath40 .
no events were observed inside the signal region and the new upper limit on the branching ratio of the decay was set to be br @xmath41 at the 90% confidence level .
the result improves the previous published limit given by the run-2 analysis by a factor of 2.6 , and the e391a experiment as a whole has improved the limit from previous experiments by a factor of 20 .
the e391a experiment was also the first step of our step - by - step approach toward the accurate measurement of the decay .
the experiment was planned to confirm our experimental approach , and this purpose was well achieved .
several points need to be noted .
first , we found solutions to several technical issues , such as the pencil beamline , differential pumping for ultra - high vacuum , low - threshold particle detection with a hermetic configuration , in situ calibration , _ etc .
_ , which can be successively used in the next step .
we encountered some technical problems that exist in the current apparatus , such as insufficient thickness and segmentation of the csi calorimeter , the structure of the cv , the limitation of the ba in an environment with higher counting rate , _ etc._. they will be improved in the next experiment . secondly , we were able to control the systematic uncertainties to be small in the estimate of the single event sensitivity , as described in the previous section .
a small systematic error is essential for an accurate determination of the branching ratio .
we are confident that the branching ratio of the decay can be measured accurately with this method .
a third point concerns the understanding and estimation of backgrounds . in the experiment for the decay , elimination of all possible backgrounds is the only effective way to identify the decay , and it can be achieved by a profound understanding of backgrounds . the dominant source of backgrounds was the result of beam interactions .
although the background from this source was considered to be as serious as those from decays in the experiment , its understanding and estimation were very difficult to assess before the work reported here .
the background mechanisms were clearly understood and divided into three different sources , cc02 , cv-@xmath42 , and cv-@xmath7 .
methods were developed to estimate them with rather small systematic errors . based on this experience , we clearly know the direction of upgrades to minimize the beam backgrounds in the next step .
the backgrounds from other decays were reduced to be negligibly small in the current experiment , mainly due to the success of applying tight vetoes with a hermetic configuration .
one of the important results is the invariant mass distribution of the events with four photons , shown in fig .
[ fig_4gmass ] .
the distribution was well reproduced by the simulations .
in particular , the low - mass region was well described by decays with a small contamination by mis - combination events of four photons from decays . the relation between and is similar to the relation between and . by taking their branching ratios into account
, it was found that the low - mass region could to be reproduced after reducing the yield by several orders of magnitude .
reproduction could not be achieved without good simulations of small signals from the detector , to which a tight veto was applied .
this achievement indicates that a similar direction will be promising in the next step .
the e391a experiment was also able to study other decay modes including s in the final state , such as @xmath43 and @xmath44 @xmath45 .
the reconstruction method and the hermetic veto system are also valid in the analysis of these modes .
results based on portions of the data have already been published @xcite , and further studies are in progress with the entire data set .
the next step , the koto experiment @xcite at the new j - parc accelerator @xcite , is now in preparation .
most of the improvements pointed out above are being implemented .
the new beamline has been constructed ; runs to evaluate properties of the beamline started in october , 2009 .
we are grateful to the crew of the kek 12-gev proton synchrotron for successful beam operation during the experiment .
we express our sincere thanks to kek staff , in particular the directors : professors h. sugawara , y. totsuka , a. suzuki , s. yamada , m. kobayashi , f. takasaki , k. nishikawa , k. nakai , and k. nakamura for their continuous encouragement and support .
we also express our sincere thanks to the many colleagues in universities , institutes , and the kaon - physics community for their continuous encouragement and support .
this work was partly supported by grant - in - aids from mext and jsps in japan , grants from nsf and doe of us , grants from nsc in taiwan , grants from krf in korea , and the istc project from cis countries .
+ @xmath46present address : laboratory of nuclear problem , joint institute for nuclear research , dubna , moscow region , 141980 russia + @xmath47present address : kek , tsukuba , ibaraki , 305 - 0801 japan .
+ @xmath48present address : cern , ch-1211 geneva 23 , switzerland .
+ @xmath49deceased + @xmath50present address : spring-8 , japan .
+ 99 see e.g. , l.s .
littenberg , j. phys .
japan * 76 * , 111006 ( 2007 ) .
see e.g. , a. r. barker and s. h. kettell , ann .
nucl . part .
sci . * 50 * , 249 ( 2000 ) .
littenberg , phys .
d * 39 * , 3322 ( 1989 ) .
see e.g. , d. bryman _
et al . _ , int .
* 21 * , 487 ( 2006 ) .
a. j. buras , s. uhling , and f. schwab , rev . mod .
phys . * 80 * , 965 ( 2008 ) , and references therein .
n. cabibbo , phys .
* 10 * , 531 ( 1963 ) ; m. kobayashi and t. maskawa , prog .
phys . * 49 * , 652 ( 1973 ) .
f. mescia and c. smith , phys .
d * 76 * , 034017 ( 2007 ) .
ckmfitter group ( j. charles _ et al .
_ ) , eur . phys .
j. c * 41 * , 1 ( 2005 ) , updated results available at : http://ckmfitter.in2p3.fr/ ; utfit collaboration ( m. bona _ et al .
_ ) , j. high energy phys .
* 07 * , 028 ( 2005 ) , updated results available at : http://www.utfit.org/. y. grossman and y. nir , phys .
b * 398 * , 163 ( 1997 ) .
j. brod and m. gorbahn , phys .
d * 78 * , 034006 ( 2008 ) . and c. smith , + http://www.lnf.infn.it/wg/vus/content/krare.html , + and reference therein .
a. alavi - harati _ et al .
_ , phys .
d * 61 * , 072006 ( 2000 ) .
j. k. ahn _
et al . _ ,
d * 74 * , 051105r ( 2006 ) ; k. sakashita , _
search for the decay _ , ph.d .
thesis , osaka university , 2006 .
j. k. ahn _
et al . _ ,
* 100 * , 201802 ( 2008 ) ; g. n. perdue , _ search for the rare decay _ , ph.d .
thesis , university of chicago , 2008 ; t. sumida , _
search for the decay _ , ph.d .
thesis , kyoto university , 2008 .
s. ajimura _
et al . _ ,
instrum . meth .
a * 435 * , 408 ( 1999 ) ; s. ajimura _
_ , nucl .
instrum .
a * 552 * , 263 ( 2005 ) .
t. inagaki _
et al . _ , kek internal report * 96 - 13 * ( 1996 )
. h. watanabe _
et al . _ ,
instrum .
a * 545 * , 542 ( 2005 ) ; h. watanabe , _ beam line for the study of the decay _ , ph.d .
thesis , saga university , 2002 .
the csi crystals used in the e162 experiment at kek - ps were recycled .
the csi crystals were borrowed from the ktev experiment at fermilab .
m. doroshenko _
et al . _ ,
instrum .
a * 545 * , 278 ( 2005 ) ; m. doroshenko , _ a measurement of the branching ratio of the decay _ , ph.d .
thesis , graduate university for advanced studies ( sokendai ) , 2005 .
t. inagaki _
et al . _ ,
instrum .
res . a * 359 * , 478 ( 1995 ) .
t. inagaki _
_ , high energy news * 23 - 1 * , 13 ( 2004 ) ( in japanese ) .
y. yoshimura _
_ , _ extrusion scintillator made of ms resin _
( to be published ) ; y. yoshimura _ et al .
_ , nucl .
instrum . meth .
res . a * 406 * , 435 ( 1998 ) , which is a report on the injection molding of scintillator made of ms resin .
et al . _ ,
instrum .
. phys . res .
a * 522 * , 477 ( 2004 ) .
y. tajima _
et al . _ ,
instrum .
a * 592 * , 261 ( 2008 ) .
t. k. ohska _
_ , kek report * 85 - 10 * , 1985 ; ieee trans .
* 33 * , 98 ( 1986 ) .
midas homepage , https://midas.psi.ch ; https://daq-plone.triumf.ca/sr/midas ; + e391a - midas , + http://www-ps.kek.jp/~e391/exp/daq/index.html . h. s. lee , _
search for the rare decay at e391a _ , ph.d .
thesis , pusan university , 2007 ; s. podolsky , _ new experimental value of the upper limit of the decay rate _ , ph.d .
thesis , b. i. stepanov institute of belorussian academy of science , 2007 ( in russian ) .
g. battistoni _
et al . _ , _ the fluka code : description and benchmarking _ , proceedings of the hadronic shower simulation workshop 2006 , fermilab 68
september 2006 , m. albrow , r. raja eds .
, aip conference proceeding 896 , 31 - 49 ( 2007 ) .
search for the rare decay _ , ph.d .
thesis , university of chicago , 2009 ; h. morii , _ search for the at the e391a experiment _ , talk at 2009 kaon international conference 2009 .
et al . _ ,
d * 76 * , 011101(r ) ( 2007 ) ; j. d. nix , _ first search for @xmath51 and @xmath52 _ , ph.d .
thesis , university of chicago , 2006 .
y. c. tung _ et al .
_ , phys .
lett . * 102 * , 051802 ( 2009 ) .
j. comfort _
et al . _ , ( j - parc e14 collaboration ) , `` proposal for experiment at j - parc '' ( 2006 ) . | the first dedicated search for the rare neutral - kaon decay has been carried out in the e391a experiment at the kek 12-gev proton synchrotron .
the final upper limit of 2.6 @xmath0 at the 90% confidence level was set on the branching ratio for the decay . |
in the first part of this supplementary material we give details of the experimental methods used .
we then more formally outline the protocol for a one - bit message . for longer messages
, the protocol can be iterated , and the security parameters readjusted to compensate for the repeated ( but independent ) use of the single bit protocol .
we then describe the unambiguous state elimination ( use ) measurement employed by the recipients , and discuss some crucial properties of the multiport which the security against repudiation relies on . after this
we state definitions of security , state some inequalities which are used in the security analysis , and then proceed to analyze security against repudiation and forging , and the robustness of the three - party no - memory qds protocol , realized using coherent states with four possible phases . in the experiments detailed in the main letter ,
the coherent states are generated by a temperature - stabilized pulsed vertical cavity surface emitting laser ( vcsel ) emitting at a wavelength of 850.17 nm and attenuated to the desired mean photon number per pulse @xmath13 using spatial interception by a knife edge connected to a computer controlled stepper motor .
the uncertainty in alice s phase encoding was primarily due to amplitude fluctuation in the electrical driving signal to the lithium niobate ( linbo@xmath31 ) phase modulator and corresponded to @xmath321.6 @xmath33 10@xmath34 radians , or @xmath320.2% of the separation between the four phases used .
the multiport exhibited a loss of 7.7 db , the receiver s beamsplitter 5.1 db and each demodulation interferometer 9.1 db .
photodetection is carried out by commercially available thick junction geiger - mode silicon single - photon avalanche diodes ( si - spads ) @xcite with a mean detection efficiency of 40.5% ( at a wavelength of 850 nm ) , dark count rate of 320 counts per second , and timing jitter of 380 ps , and time stamping electronics that could record photon arrival times at time intervals of 1 ps , but exhibited 12 ps independent timing jitter @xcite .
although using an emission wavelength near 1300 @xmath35 m or 1550 @xmath35 m would permit compatibility with standard telecommunications optical fiber , detector technologies for these wavelengths @xcite are not as advanced as those at visible and shorter near infra - red wavelengths , and suffer from higher dark count and afterpulsing rates @xcite which would increase error rates .
although the dark count - rate of the detectors is relatively small and the possibility of intersymbol interference low , we time - gate the raw detector events using a window of duration @xmath321 ns centered on the expect photon arrival time to temporally filter spurious counts from the recorded events . at a delay of 10 ns after the detection peak the probability of a photo - generated count was 2 @xmath33 10@xmath36 @xcite . the time - gated count rate is approximately 91% of the raw count rate .
phase is a relative measurement and a reference for zero phase must be provided in some manner . in our system
we employ a technique frequently employed in quantum key distribution systems which use phase - encoded states @xcite .
an asymmetric unbalanced mach - zehnder interferometer is used to provide a time - delayed reference pulse with zero phase , which propagates through the same fiber as the phase - modulated signal " .
the phase reference pulse is delayed by 5 ns relative to the signature state pulse at alice , and the delay is canceled in each receiver .
the phase reference and signature state pulses have orthogonal polarizations and therefore can be correctly routed by the receivers using a polarization beamsplitter ( pbs ) @xcite .
all fiber components are constructed from polarization - maintaining
panda eye " fiber @xcite with a core diameter of 4.4 @xmath35 m .
polarization routing increases the time - gated detector count - rate of our system from 43% of the raw count - rate as observed using a 50:50 beamsplitter to 91% of the raw count - rate when using a pbs .
we comment further on establishing a common phase reference below in section iv , in connection with the theoretical description of the use measurement .
changes in temperature or mechanically induced stress will result in changes in the relative path - lengths of the interferometers in the system .
these changes in path - length lead to changes in the visibility and , typically , to increased errors .
the recipients employ variable delay air - gaps consisting of a fixed collimating launch lens and a collection lens connected to a piezo - electric computer - controlled linear actuator to maximize the fringe visibility in the multiport and the demodulation interferometers @xcite .
the piezo - electric linear actuators had a step size of approximately 15 nm and a maximum travel of approximately 1.5 @xmath35 m .
an adjustable position knife edge was placed in the collimated beam in each air - gap to match the loss in each optical path .
a mutually known tuning signal is used to establish maximum fringe visibility prior to signature transmission . for this demonstration a shared 10
mhz rb clock was used to provide a reference for both sender and receivers , but it is possible to use separate clocks with connections to the signals broadcast by the global positioning system ( gps ) satellites to maintain a shared time reference @xcite . * _ distribution stage _ 1 . for each possible
future message @xmath37 , alice generates two copies of a sequence of coherent states ( called _ quantum signatures _ ) @xmath38 where @xmath39 , @xmath3 is a real positive amplitude , @xmath40 are randomly chosen , and @xmath0 is a suitably chosen integer .
the state @xmath41 and the sequence of signs @xmath42 are called the _ quantum signature _ and the _ private key _ , respectively , for message @xmath43 . the individual state @xmath44 we call the @xmath45 _ quantum signature element _ state for message @xmath43 .
alice sends one copy of @xmath41 to bob and one to charlie , for each possible message @xmath46 and @xmath47 .
[ prot : usd ] bob and charlie send their sequences @xmath41 for @xmath46 and @xmath47 , one signature element at a time , through the qds multiport .
for each signature element they ( a ) note whether photons are registered at their multiport null - port .
they also ( b ) measure the multiport signal states using the use measurement for @xmath48 ( see below ) .
for every element of each quantum signature ( for @xmath37 ) , they store which detectors detected photons ; each detector rules out one possible phase state .
they therefore store sets of six numbers ( hexaplets ) of the form @xmath49 , where @xmath50 and @xmath51 . here
@xmath52 means that no photons were detected at the @xmath53 detector ( that is , the phase @xmath54 is not ruled out ) , while @xmath55 means that there was at least one photon detected at the @xmath53 detector ( that is , the phase @xmath54 is ruled out for this element ) .
note that by @xmath53 detector we symbolize the `` not @xmath54 '' detector .
* _ messaging stage _ 1 . to send a signed one - bit message @xmath56
, alice sends @xmath57 to the desired recipient ( say bob ) .
2 . bob checks whether @xmath57 matches his stored sequence .
in particular , he counts the number of elements of @xmath58 which disagree with his stored hexaplets .
therefore , for a given element @xmath59 of the signature , if alice s declaration was @xmath54 , bob needs to check if @xmath60 is 0 or 1 . if @xmath55 , he registers one mismatch . in other words ,
a mismatch is registered whenever alice s declaration for a given element has been eliminated by bob s use measurement .
bob checks whether the number of mismatches is below @xmath61 , where @xmath14 is an * authentication threshold*. 3 .
provided the authentication threshold was not breached , before accepting the message , bob checks that he has no reason to abort the protocol . if the number of signature elements for which non - zero null - port counts are registered is higher than a threshold @xmath62 for some @xmath63 he aborts .
if the authentication threshold was not breached , and the protocol has not been aborted , bob accepts the message coming from alice .
4 . to forward the message to charlie , bob forwards to charlie the pair @xmath64 he received from alice .
charlie tests for mismatches similarly to bob , but to protect against repudiation by alice , he uses a different threshold .
charlie checks if the number of mismatches is below @xmath65 where @xmath15 is the * verification threshold * , with @xmath66 .
5 . for charlie to accept the forwarded message , provided the verification threshold was not breached , he confirms that he has no reason to abort the protocol by checking the null - port counts in the same way as bob . in quantum communication
it is important to be able to distinguish different signal states from each other .
signal states may be non - orthogonal , either on purpose , for example in protocols for quantum key distribution , or because of unavoidable noise .
quantum measurements aiming to distinguish between quantum states can be optimized in different ways , for example by minimizing the probability to select the wrong state ( minimum - error measurement ) , by minimizing the average cost of selecting a measurement result ( minimum - cost measurement ) , by maximizing the classical information in the result about what state was sent ( maximum mutual information measurement ) or by requiring that an obtained result is always correct , but the measurement is allowed to sometimes give an inconclusive answer .
the last type of measurement is known as unambiguous state discrimination ( usd ) . in all cases ,
these measurements aim to determine what state was sent , and are therefore called `` quantum state discrimination '' measurements .
one can also instead consider `` quantum state elimination '' or
quantum state exclusion " , which in a sense is a generalization of state discrimination @xcite .
the difference is that one no longer aims to identify which state was sent , but instead which state(s ) were _ not _ sent , i.e. to rule out or eliminate one or more of the possible states . if one eliminates @xmath67 states , this is the same as identifying the state that was sent , and it is in this sense that state discrimination is a special case of state elimination . just as for quantum state discrimination
, one can for example aim for minimum - error state elimination , minimum - cost state elimination or unambiguous state elimination ( use ) . in @xcite , caves _ et al .
_ discussed the compatibility of quantum - state assignments , and this can be seen as quantum state elimination .
the case of excluding one of three states was also explicitly given .
in @xcite , barnett discussed unambiguous quantum state elimination of a single state out of many possible ones , and was the first to use the term .
more recently , this type of measurement was termed quantum state exclusion " in @xcite . in our protocol ,
bob and charlie perform unambiguous quantum state elimination .
the advantage of using unambiguous measurements is that whenever bob or charlie obtains a result other than an inconclusive result , they can be certain that this result is correct ( in the ideal case ) .
bob does not know for which of the elements of the signature charlie has obtained unambiguous outcomes ( or what the outcomes are ) . in order to forge
, he needs to make a guess for _ all _ the elements of the signature .
bob s best forging strategy will therefore be some type of minimum - cost state discrimination measurement . the advantage of using state _ elimination _ instead of state _
discrimination _ is that the success probability for eliminating fewer than @xmath67 out of @xmath68 states can be ( much ) higher than the success probability of full usd , especially when considering experimental imperfections . also , even if one aims to perform a full usd measurement , when it fails to unambiguously identify the state it may sometimes still rule out some of the states .
this is the case for the best known linear optical realization of a usd measurement among symmetric coherent states , suggested by @xcite and realized for four coherent states in @xcite . for an application such as quantum digital signatures ,
it is advantageous for bob and charlie to use all available information of what states have been ruled out .
this is what led us to employ a use rather than usd measurement in our realization of the qds protocol . in the current qds protocol
, alice chooses from four coherent states @xmath69 . the use measurement results in eliminating one , two or three possible states ( unambiguous or conclusive result ) or none at all ( ambiguous or inconclusive result ) . for each signature element
, bob directs the beam received from alice onto a 50/50 beam splitter , and charlie does the same .
one of the resulting beams is interfered on another 50/50 beam splitter with a reference beam in the state @xmath70 , and the other one is interfered with a reference beam in the state @xmath71 , as shown in figure [ fig : use ] .
note that the reference beam is obtained from a reference pulse sent by alice , in order to ensure that the outgoing states are as expected on recombination at beam splitters 2 and 3 ( see below ) . here
we assume that the reference pulse is not tampered with .
however , it is worth noting that intuitively one can not obtain further repudiation and forging possibilities through tampering with the reference pulse .
repudiation is guaranteed by the symmetrization of the states due to the multiport , and this does not depend on the timing of the reference pulse .
as far as forging is concerned , tampering with the reference pulse can not help bob to obtain further information about the signature , and since the original phase sent by alice is unknown to him , he can not use the reference pulse to steer the outcome of charlie s use measurement towards his declaration either . in principle , the correct phase reference states for the use measurement could also be obtained without a reference pulse sent from alice to bob and charlie .
if the parties instead have synchronized clocks , then bob and charlie could themselves prepare the required use reference states .
given the above considerations , we can now see that the resulting output state , assuming that the incident state was @xmath72 , and assuming no losses , is @xmath73 where the four output modes refer to either bob s or charlie s output modes . from this
it is clear that if the state that alice sent was @xmath74 to start with , then the first mode will contain the vacuum state .
therefore , if any photons are detected in this mode , it follows that the state alice sent can not have been @xmath74 .
similarly , if alice sent the state @xmath75 , then the second mode necessarily contains the vacuum state , and therefore detecting any photons here unambiguously rules out the state @xmath75 .
seeing any photons in output modes 3 and 4 rules out the states @xmath76 and @xmath77 , respectively . to summarize , any detector clicking rules out one possible state .
if three detectors click we have fully and unambiguously ( in the ideal case ) determined which state alice sent . in other cases ,
bob or charlie may only have eliminated some of the possible states . for every signature element , bob and charlie register which detectors ( if any ) click . in the messaging phase , when alice ( or a forger ) declares what a particular signature element state was , bob and charlie check if this state was ruled out by their elimination measurement .
if it was , they register a mismatch .
losses in the setup will not affect the conclusive nature of the outcomes , only the success probability .
dark counts and other imperfections , on the other hand , may lead to errors in conclusive outcomes .
this will be discussed later in connection with experimental results .
we should also note that the use measurement described above may not be the optimum use measurement .
exploring the properties of optimal use measurements is an interesting problem . nevertheless , the great advantage is that the realization we have described requires only linear optics and commercially available photodetectors . the multiport is a passive linear optical device with four input modes and four output modes , comprising four 50/50 beamsplitters .
two of the input modes always contain the vacuum state .
the beam splitters act on the field operators according to the multiport and the input - output relations for coherent state inputs are illustrated in figure [ fig : mult1 ] .
the top two beam splitters are held by bob and the lower two by charlie . since two of the four input modes of the multiport
are always set to the vacuum state by bob and charlie , we will refer to the remaining two modes as the input modes of the multiport . for an input state where two of the input modes are in the vacuum state
while the other two are coherent states @xmath3 and @xmath78 , the multiport acts according to alice controls the ( non - vacuum ) input states of bob and charlie . as proven in detail in @xcite , it follows that the reduced state in bob s signal port and the reduced state in charlie s signal port are identical , for each individual signature element , even if the initial state sent by alice was a general , possibly entangled state .
that is , the reduced states for each signature element .
this holds even if alice employs any type of entanglement , such as entanglement between different signature elements , entanglement between bob s and charlie s states , or entanglement with some other state which could be retained by alice .
one can further prove a somewhat stronger property , that the resulting state for bob and charlie , for one signature element position , is symmetric under swap of bob s and charlie s subsystems .
the presented quantum digital signature protocol is designed to be immune to two types of malicious activities : forging and repudiation .
immunity to forging means that any receiving party will with high probability reject any message which was not sent by alice herself .
immunity to repudiation guarantees that if bob accepts a message from alice , then with high probability the same message will pass verification with charlie as well .
that is , alice can not make bob and charlie disagree on the validity ( and consequently the content ) of her message .
more formally we have the following : * we say that a protocol realizing qds is * secure against forging * if the probability of a recipient successfully producing , without receiving it from alice , a private key for a message @xmath56 , which will pass verification by the other recipients , is _ decaying exponentially quickly as a function of the quantum signature length _ @xmath0 . *
we say that a protocol realizing qds is * secure against repudiation * if , for any malicious activity by alice , the probability of a message failing verification with one recipient once it has already passed authentication with another recipient is _ decaying exponentially quickly as a function of the quantum signature length _ @xmath0 .
* we say that a protocol realizing qds is * robust * if , when all parties are honest , a message will be authenticated and verified except with a probability _ decaying exponentially quickly as a function of the quantum signature length _ @xmath0 . defining the level of security for a qds scheme is complicated .
unlike for qkd , it is not fixed which participants are honest and which are malevolent .
in other words , some choices of parameters may lead to a certain level of guaranteed security against forging and a different level for security against repudiation .
however , one can choose @xmath14 and @xmath15 such that the resulting probabilities for any undesirable event ( repudiation , forging or honest rejection ) are equal . in this case
, we will call the probability for an undesirable event the _ security level _ of the qds scheme . in the last section
, we will comment further on this , and on the resulting rate and length @xmath0 required to sign a half - bit message for a given security level . both for repudiation and for forging one can distinguish different types of malicious attacks . in _
individual _ attacks , the cheating party employs a strategy separately and independently for each signature element . in _
collective _ attacks , there may be classical correlations between strategies for different signature elements .
these attacks can also be called _
separable_. _ coherent _ attacks are the most general type ; here a cheating party can employ any type of correlations , including entanglement and measurements in an entangled basis . for forging , we can additionally distinguish between _ passive _ and _ active _ attacks , depending on whether a forger acts maliciously only during the messaging stage , or also during the distribution stage .
security is proven for all types of repudiation attacks , and all types of forging except coherent forging attacks .
we will treat individual repudiation and forging in more detail .
the analysis for other types of attacks follows the treatment in @xcite , and we will only point out the differences . intuitively , the protocol is secure also against coherent forging , but a full analysis of this remains an important task and is the subject of ongoing work .
let @xmath79 be independent random variables , each attaining values @xmath80 or @xmath81 .
let @xmath82 be the empirical mean of the variables , and let @xmath83 be the expected value of @xmath84 .
then for all @xmath85 we have in a repudiation event , alice manipulates the quantum signature states which are sent in the distribution stage , in such a way that during the messaging stage , the same declaration will be confirmed by one party ( checking against the threshold @xmath14 ) but rejected when forwarded to another party ( who checks against the threshold @xmath15 ) .
repudiation is successful , if ( say ) bob authenticates ( @xmath86 ) , charlie rejects ( @xmath87 ) and there is no abort ( @xmath88 ) . to bound this probability
, we can consider the minimum of the following two probabilities , bob authenticating , and charlie rejecting to verify , i.e. there are two things to note to understand intuitively why such a bound would be useful .
first , we should note that due to the symmetricity of the multiport @xmath89 and @xmath90 are _ not _ independent .
second , we should note that the threshold for bob authenticating is lower than that of charlie verifying . here
we will consider individual strategies by alice .
general separable strategies and general coherent strategies , follow by reduction to individual strategies as in @xcite .
therefore the best general strategy for alice is given by individual attacks . in this type of attack
alice sends quantum signatures , possibly different , to bob and charlie .
each pair of bob s and charlie s corresponding quantum signature elements are not correlated with other pairs ( classically or through entanglement ) , but are not necessarily identical , pair to pair .
that is , the global state of the quantum signature is in a factorized form , with respect to the partition in different signature element pairs .
there may be entanglement within each bob - charlie signature element pair . because bob and charlie pass their quantum signatures through the multiport , the outbound quantum signature element at the signal ports of bob and charlie is always symmetric under swap of bob and charlie .
that is , the quantum signatures bob and charlie collect at their multiport signal ports are such that the reduced density matrix for each element of the signature of bob is the same as charlie s . for each signature element , alice s declaration agrees with bob s ( charlie s ) measured eliminations with probability @xmath91 if the declaration is not eliminated , and disagrees with probability @xmath92 if it is eliminated ( mismatch ) . those probabilities , provided the multiport has been appropriately adjusted and calibrated , are identical for bob and charlie .
we can define the _ average _ probabilities @xmath93 , with @xmath94 .
furthermore , we define as @xmath95 the empirical ( observed ) percentage of mismatches over the total number ( @xmath0 ) of elements of the signature .
it is easy to see that bob authenticating means that @xmath96 , which using eq .
( [ hoeffding 2 ] ) leads to provided that @xmath97 .
note that @xmath98 is an arbitrarily small positive number .
we note that if this is satisfied , the probability decays exponentially .
charlie failing to verify means that @xmath99 which using eq .
( [ hoeffding 1 ] ) leads to provided that @xmath100 .
alice s only freedom , in her attempt to repudiate , is to choose different @xmath101 s .
in reality , alice does not have full freedom for these choices ( as the povm elements of bob and charlie s measurements do not have orthogonal supports ) , but since we are looking for a bound , we may assume a worst - case setting and assume that she does have full freedom .
noting that @xmath102 , alice can only choose the optimum for repudiation value of @xmath103 .
therefore , provided that we choose the probability for repudiation decays exponentially for _ all _ choices of @xmath103 . from eq .
( [ repudiation 1 ] ) , we can see that alice s best strategy to achieve repudiation is to choose @xmath103 in such a way that the minimum of ba and cr is the greatest . from eq s ( [ ba]-[cr ] )
we get that the optimum value is for @xmath104 .
this gives the following bound for repudiation under individual attacks , it is interesting to note that there is an optimal value of @xmath103 from alice s point of view , and that her repudiation probability depends on this .
therefore , for the given bound we consider , it is no better for her to choose different values for different @xmath92 , and thus a strategy where alice sends identical states for each signature element is at least as effective as full individual strategy for alice . as also noted in the main text , forging is defined as when a cheating party ( say , bob ) convinces an honest party ( charlie ) that alice had sent a classical message @xmath43 , when alice has sent no message or another message .
to do so , the malevolent party needs to guess alice s declaration . throughout this section ,
when we mention declaration , and matching and mismatching with the declaration , this is the `` fake '' declaration of bob , which is made after bob uses all his resources to make the best possible guess .
note also that security against forging as defined also ensures security against message tampering . at the end of the distribution stage bob / charlie makes a use measurement .
the outcome of the measurement is that they rule out one or more of the possible phases .
these outcomes should give a guaranteed minimum advantage to anyone who declares the actual phase sent , against anyone who declares some other phase .
if this is the case , then there are choices of @xmath27 and @xmath0 that would make the protocol secure , as we will see below .
there are two different characterizations of forging attacks .
the first concerns whether or not the malevolent party was dishonest from the beginning or only during the messaging stage .
_ passive forging _ is where bob is honest during the distribution stage and simply attempts to guess the signature using his own copy , in such a way that his declaration gets accepted by charlie . _
active forging _ is where bob is dishonest also during the distribution stage , where he attempts to modify charlie s quantum signature so that it helps him subsequently during the messaging stage to make charlie agree with his own declaration . here
we will first examine passive forging in greater detail , and then argue that active forging can be reduced to modified passive forging attacks by placing a limit on the multiport null - port counts during the distribution stage .
the second characterization concerns the type of measurements and actions that the malevolent party can do , similar to as for repudiation .
there are three types of attacks , ( 1 ) individual attacks , where bob is restricted to individually measuring the elements of his signature and also to tampering with the response states individually , ( 2 ) collective attacks , where bob is allowed to include classical correlations between the measurements and response states for different signature elements and ( 3 ) coherent attacks , where bob is allowed to make any possible joint measurement and also send any possible entangled state to charlie . in this paper
, we will only examine the first two types of attacks and leave coherent forging strategies for future work . from eq .
( [ condition 1 ] ) it is clear that it is easier to attempt to forge a message in the verification stage rather than in the authentication stage . in other words , it is easier to convince charlie that bob forwards the message received from alice , than charlie convincing bob that he directly sends the message ( pretending to be alice ) .
we have defined as successful forging the scenario where charlie verifies the fake message sent by bob and there was no abort .
abort occurs only when the null - port of the multiport register above than @xmath62 .
when alice is honest , abort may thus occur only if bob attempts to tamper with the signature at the distribution stage , that is , for active forging . in passive attacks bob
does not interfere during the distribution of the signatures , and therefore charlie obtains as quantum signatures what alice , who is honest , sends .
bob s strategies comprise all the possible measurements he could perform on his copy of the quantum signature , resulting in a declaration for each signature element .
bob aims to minimize the probability that his declarations will lead to a rejection , when compared with the eliminations that charlie has obtained from his use measurements .
bob knows what measurement charlie , who is honest , will make on the signature charlie receives from alice . in particular , bob knows what charlie s measurement statistics is , in what ways charlie s measurement may be imperfect , and also what charlie s verification threshold is , and bob can adjust his measurement accordingly .
bob s optimal strategy will be a measurement which maximizes the probability of charlie accepting bob s declaration as a whole .
recall , charlie accepts if the number of incorrectly declared phases is below a certain number , determined by the verification threshold .
the optimal measurement for this problem is a so - called minimum cost measurement @xcite . in particular ,
any declaration of bob , as a whole , for all signature elements , has a cost equal to the probability that charlie will reject this declaration .
this means that bob s goal is _ not _ in general to minimize the probability of a mismatch for each individual element .
bob goal is instead to meet the verification threshold . to take a simple example , suppose that there are two signature elements , and that charlie will accept the message if there is at most one mismatch
bob should then concentrate on avoiding two mismatches , and declarations with zero and one mismatches both carry no cost .
if bob is limited to collective forging strategies , then it is in bob s favor to minimize the probability of a mismatch for each individual element , as we will now argue .
that is , it is optimal to make independent minimum - cost measurements for each signature element .
collective strategies would allow bob to change the measurement he makes on subsequent signature elements , conditioned on the outcomes he obtained for previous elements .
there is however no correlation between the phases of different signature elements , as alice chooses these independently of each other .
charlie s eliminations also occur independently for each element of the signature .
as long as bob is not allowed to make a measurement in an entangled basis , any knowledge of the phase of one element does not alter his chances of guessing another element correctly , and does not alter what measurement he should make on this signature element .
therefore his best strategy is to make a minimum - cost measurement for each individual signature element . in other words ,
bob can not benefit from classical correlations between his measurements for different signature elements .
therefore , provided that one restricts attention to collective attacks , bob s best strategy is to make a minimum - cost measurement for each signature element .
we proceed to bound the probability of success for this type of individual forging attack .
we will prove that the forging probability is less than the minimum - error probability multiplied by the guaranteed advantage that an honest participant has .
for a given individual signature element , we define the cost matrix as a matrix where the rows corresponds to which state alice sent ( @xmath105 ) , while the columns correspond to the detectors @xmath106 .
each matrix element @xmath107 can be taken equal to the probability that if the @xmath108th state is sent , then charlie s @xmath109th detector clicks .
this is because bob should avoid declaring a phase that charlie has eliminated .
his cost for making a particular declaration will therefore be proportional to the probability that charlie has ruled out this state .
as we have mentioned earlier , in the ideal case , charlie should never rule out the phase that alice sends , and thus in the ideal case , bob s cost matrix would have zeroes on the diagonal . however
, due to losses and noise ( mainly due to the multiport ) , we do have the following actual experimental cost matrix , for the case where @xmath19 , in the experiment , we considered several different values for @xmath13 ranging form @xmath19 to @xmath110 .
we chose to give the cost matrix for @xmath19 since , using the bounds we have , this gives the best performance ( requires a smaller length @xmath0 for given security level ) among the tested values of @xmath3 .
the minimum guaranteed advantage is the smallest difference between a diagonal element and an off - diagonal element in the same row .
the reason for this will become apparent soon . for our case the minimum advantage is @xmath111 , as is seen from the fourth row . in the general case
, one has to assume that bob has full knowledge of charlie s actual measurement statistics , and thus this type of cost matrix should be used in a security analysis of an actual realization .
we recall that @xmath95 is the empirically observed percentage of mismatches .
bob succeeds in forging , if charlie , after receiving bob s ( fake ) declaration , finds less than @xmath65 mismatches ( @xmath112 ) , and thus verifies . consequently
, bob tries to guess the correct signature by making a minimum - cost measurement for each element signature , i.e. the measurement that , given the cost matrix in eq .
( [ exp cost matrix ] ) , minimizes the overall cost given by here @xmath113 is the prior probability of each state ( alice sends each of the for phases with equal probability ) , and @xmath114 are the elements of bob s measurement , a probability operator measure ( pom ) , also called a positive operator - valued measure ( povm ) .
the minimum cost @xmath115 is the minimum possible probability that bob s declaration for a single signature element has been eliminated by charlie .
we will give a bound for this cost , for our actual experimental realization , below .
for now , we note that for each signature element we have a binary random variable that takes value `` @xmath81 '' ( match ) or `` @xmath116 '' ( mismatch ) , and @xmath115 is the expected probability for a mismatch .
since we assume that each of the elements is measured separately , the random variables are independent and the hoeffding inequalities hold . using eq .
( [ hoeffding 2 ] ) the probability of forging ( which occurs if the observed mismatches are fewer than the threshold @xmath117 ) is bounded as provided that we choose @xmath15 such that @xmath118 .
if this holds , we see that the forging probability decays exponentially in the signature length @xmath0 . before proceeding to derive a bound on the possible @xmath115 that bob
can achieve , we should point out that it is crucial that there is a gap between the minimum probability of a forger to have his declaration eliminated , @xmath115 , and the probability @xmath119 of the correct declaration being eliminated if all parties are honest . defining this gap as @xmath120 , the choices of @xmath14 and @xmath15 for a robust and secure protocol should be within this gap ( we elaborate more on parameter choices at the end ) .
we proceed to bounding the minimum cost of the measurement that a malevolent bob performs .
methods for bounding the minimum cost and properties of the minimum - cost measurements that are used here can be found in @xcite . since we trying to find a bound on the forging probability , we assume the best case scenario for bob , and therefore we are looking for a _ lower _ bound on the minimum cost
. we will use three properties of minimum - cost measurements . to start with ,
the minimum cost for a given cost matrix @xmath107 can always be bounded by the cost of any other strictly smaller cost matrix @xmath121 for all @xmath122 .
secondly , define a _ constant - row _ matrix as a matrix of the form if we add or subtract such a constant - row matrix from any given cost matrix , then the measurement that minimizes the cost for the new cost matrix does not change , while the minimum cost simply shifts by a constant @xcite .
it is easy to see that the average cost for any measurement , not just an optimal one , shifts by @xmath123 .
intuitively , for a constant - row cost matrix , no matter what measurement one makes , the cost of all outcomes ( i.e. the matrix elements in a given row ) are equal .
the cost does not depend on what outcome is obtained , but only on which state alice sends , which in its turn depends on the prior probabilities @xmath124 .
the third property that we will use is the fact that the minimum - cost measurement for a cost matrix of _ error - type _
, for the case of symmetric pure states , is the square root measurement ( srm ) .
a cost matrix is of error - type if all the diagonal elements are zero , and all the off - diagonal elements are positive and equal with each other .
we call this error - type , because for such a cost matrix , guessing correctly has no cost ( the diagonal elements ) , while any mistake carries an equal cost ( the off - diagonal elements ) .
the cost matrix for a minimum - error measurement , which results in the minimum probability of error @xmath125 , is of error - type , with off - diagonal elements having cost 1 .
this means that the minimum cost of any error - type cost matrix is proportional to the minimum - error probability @xmath125 . to bound @xmath115 for the cost matrix in eq .
( [ exp cost matrix ] ) , we first define @xmath126 , a constant - row matrix for which all elements in every row are equal to the diagonal elements of the matrix @xmath107 .
then we define a cost matrix @xmath127 , which has the same minimum - cost measurement as @xmath107 , but with the minimum cost shifted by @xmath128 .
finally we define another cost matrix @xmath129 which is strictly smaller than @xmath130 for all @xmath122 , such that @xmath131 for @xmath132 , and with zeroes on the diagonal .
this final cost matrix @xmath129 is of error - type , and the corresponding minimum - cost measurement is therefore the srm and the minimum cost is proportional to @xmath125 . in our case , denoting the relevant minimum costs with @xmath133 and @xmath134 , we have from eq .
( [ exp cost matrix ] ) we obtain and also @xmath135 .
this is the cost for the honest scenario , in other words , the probability that charlie has eliminated the state that alice actually sent , so that there would be a mismatch even if all parties including bob are honest , and bob simply forwards alice s correct declaration . from the matrix @xmath136
we can see that the bound on the advantage of a correct declaration compared to a wrong declaration , i.e. the guaranteed advantage , is @xmath137 .
the bound on the gap is then given by @xmath138 we compute @xmath125 using the srm @xcite , which for @xmath139 gives @xmath140 .
we obtain @xmath141 , which provides a bound on the gap @xmath18 . to confirm that the bound on the minimum cost we obtained is relatively tight
, we can obtain an upper bound for the minimum cost , along similar lines as the lower bound , which gives @xmath142 .
we therefore obtain and for active individual and collective forging attacks , the security analysis follows the treatment in @xcite . we will discuss and give the expressions for attacks where bob behaves independently and identically for each quantum signature element ( iid forging attacks )
however , independent , non - identically distributed strategies ( inid attacks ) , and also collective forging strategies , reduce to , and are not more powerful than , iid forging strategies . here
we refer the reader to @xcite .
in essence , active forging can be limited in the following way .
both bob and charlie keep track of the number of detections at the null - port of the multiport .
if one of the parties receiving the signature states ( say bob ) is dishonest during the distribution phase , the state he sends may not be identical to the state that alice sends , and if so there will on average be some detections at the null - port of the multiport .
since the protocol is aborted if the detections on the null - port exceed some number @xmath62 , we can bound how much bob can tamper with the signature states .
this effectively reduces active forging to passive forging .
then the whole security analysis follows in a similar fashion as for passive forging , except that the verification threshold @xmath15 is effectively lowered by a function of @xmath143 which is the abort threshold . by choosing @xmath143 suitably small
, we can make sure that the effective threshold @xmath144 is still greater than @xmath14 so that the protocol remains exponentially secure .
_ iid attacks_. in active strategies , bob is allowed to alter the quantum signature sent to charlie during the distribution stage , through tampering with the part of the signature state which he sends to charlie . by doing so ,
bob tries to increase his probability to later forge a message . in short , if bob alters the state in the multiport , then charlie s null - port of the multiport will occasionally ( or more often ) detect photons . by demanding that the protocol is aborted
if too many photons are detected at the multiport null - port , charlie limits the degree to which bob can manipulate charlie s quantum signature state .
the amount that this happens is bounded by using the relationship of fidelity with the trace distance as in @xcite .
this reduces bob s active attacks to ( modified ) passive forging attacks .
the forging probability is then bounded by where @xmath145 is an arbitrarily positive constant , chosen when designing the protocol , and @xmath146 is a lower bound for the minimum cost that bob can achieve when making his best guess .
@xmath146 is calculated by allowing bob to base his guess on a measurement on coherent states with amplitude @xmath147 the minimum cost for the optimal measurement @xmath148 is smaller than @xmath115 , since bob now has access to a state with greater amplitude , but can be bounded using the same method as we previously used to bound @xmath115 .
the above expression holds , and it decays exponentially , provided that after having analyzed the security of the protocol against repudiation and forging , we examine the robustness of the protocol .
if all participants are honest , it should succeed with probability approaching one exponentially quickly as @xmath0 increases . here
we will first consider the probability of rejecting a message if all parties are honest and then the probability of aborting the protocol if all parties are honest .
both of them should , and do , decay exponentially .
if all parties are honest , then the probability that the phase @xmath149 is eliminated , when the state @xmath150 with phase @xmath149 is sent , is equal to the number of detections at the detector @xmath151 over the total number of pulses of @xmath150 . this probability is the diagonal element @xmath152 of the cost matrix in eq .
( [ exp cost matrix ] ) .
therefore , averaged over all possible states , the probability that the correct state is ruled out is @xmath153 . for bob to fail to authenticate
, we need that there are more than @xmath61 mismatches .
using hoeffding s inequalities , this occurs with probability provided that @xmath154 , which can be made to decay exponentially in @xmath0 . on the other hand , abort can only occur if the number of the null - port counts exceed @xmath62 .
it is interesting to note , that if we assume that bob and charlie are honest during the distribution stage , there is no need to keep count of the counts of the null - port and no need to abort the protocol .
it is only due to active forging and the need to restrict the ability of bob to tamper with the response state , that we require to consider aborting . in order for our protocol to be correct ,
we need to choose @xmath143 to be larger , by at least some finite amount @xmath145 , than the average number of null - port counts ( dark - counts ) when all parties are honest , which is @xmath155 .
then the probability for abort if all parties are honest decays exponentially with the signature length , here we summarize the performance of the protocol .
in particular , if we consider only passive forging , we have three main inequalities bounding the probabilities for undesired events , namely together with the conditions on the parameters , from this equation and the definition of the gap @xmath120 , we can see that for positive gap , there exist choices of @xmath14 and @xmath15 that would make the protocol secure and robust . from eq .
( [ gap - adv ] ) it follows that the existence of a positive gap is guaranteed if the cost matrix of the experiment leads to a positive guaranteed advantage @xmath156 .
note also , that as stated above , we do not need to keep track of the multiport null - port counts , leading to possible abort also if all participants are honest , unless we are also considering active forging . nevertheless , according to , the probability for abort if all participants are honest also decays exponentially in @xmath0 .
we assume that we are equally interested in robustness , security against repudiation and security against forging . by choosing @xmath157 and @xmath158 ,
the bounds for the forging , repudiation and honest rejection probability become equal . since @xmath120 ,
all three equations lead to we see that with this choice of the parameters @xmath14 and @xmath15 , we achieve exponential security against forging and repudiation , and exponentially small probability of abort if all parties are honest .
the exponential decay rate depends on @xmath18 , which for the demonstration provided in the main paper is found to be small .
there are two figures of merit by which a qds scheme is evaluated .
the first one is the length @xmath0 of the quantum signature that alice needs to send to be able to securely sign a `` half - bit '' message .
the length @xmath0 is the qds analogue of the ratio of raw key length to established secure key length in qkd .
the second figure of merit is the rate by which signatures are distributed .
this is the number of bits that alice can sign per second , and is the clock rate divided by @xmath0 . for both of the above figures of merit
, one needs to define what is meant by secure and robust , i.e. how small probability of repudiation , forging and honest rejection we desire .
statements about the performance of a qds scheme should include the security level requested . in the estimates given below
, we will require that the qds scheme has a security level of @xmath23 , which means that the probability of forging , repudiation or honest rejection are all less than @xmath159 .
here we should also remark on how the length @xmath0 for a given security level depends on the losses in an experiment .
assume that we have a cost matrix coming from an experiment with given losses @xmath59 , and that we want to extrapolate the cost matrix if the losses change to @xmath160 .
since the cost matrix elements are determined by the ratio of the number of detection events to the average number of pulses sent , having higher losses implies that the overall matrix is multiplied by @xmath161 .
if the guaranteed advantage of the experiment with losses @xmath59 is @xmath162 , then the guaranteed advantage with losses @xmath160 will simply be @xmath163 . from eq .
( [ gap - adv ] ) we can therefore see that the length @xmath0 required for a given security level increases quadratically as the transmitted signal ( one minus losses ) decreases . in our experiment
, we obtained a gap @xmath22 which for a security level of @xmath23 leads to a length @xmath24 required to sign a `` half - bit '' .
this is clearly a very large number and to make a qds scheme practical , one needs to improve it significantly . however , there are three changes that can be implemented that would indeed result in a significant improvement .
we present them below along with our estimate of the expected performance of the corresponding experiments .
the first change is to use a different amplitude @xmath3 .
the smaller the @xmath3 is , the greater the value of @xmath125 , which in its turn affects the value of the gap . on the other hand , the guaranteed advantage decreases with smaller @xmath3 .
it is likely that there is an optimal value for @xmath3 , which is not in the range explored in our experiment . using data from a subsequent experiment further exploring unambiguous state elimination , for which we have a greater range of @xmath3 , and extrapolating using the known losses from the multiport
, we have good indication that the optimal value of @xmath164 is around @xmath28 .
moreover , we estimate that this would lead to a significantly larger gap of @xmath165 .
this gives a length @xmath166 for a security level of @xmath23 .
the second potential improvement is to use a protocol that does not require the multiport , which is a major source of losses . removing the multiport means that the security against repudiation needs to be guaranteed in an alternative way .
we have recently proposed some qds schemes that do not require the multiport @xcite and one could modify those protocols to use phase - encoded coherent states , similar to the current realization .
we estimate that implementing all these change will result in a gap of @xmath29 and length @xmath30 for a security level of @xmath23 .
finally , we could increase the clock rate , and therefore the transmission rate .
the phase modulators , vcsel , and driving electronics are capable of clock - rates up to 3.3 ghz .
this would not change the length @xmath0 , but it would reduce the required time to distribute the same number of secure signatures bits by a factor of 33 , and therefore significantly improve the rate . in particular , we expect that making these changes will lead to distributing signatures at a rate capable of signing 1.4 bits per second with a security level of @xmath23 .
l. widmer , _ cerberis : high - speed encryption with quantum cryptography _ in _ internet technical development and applications _ , editors e. tkacz and a kapczynski ( springer , berlin , 2009 ) , advances in intelligent and soft computing * 64 * , 217 . | digital signatures are widely used to provide security for electronic communications , for example in financial transactions and electronic mail .
currently used classical digital signature schemes , however , only offer security relying on unproven computational assumptions .
in contrast , quantum digital signatures offer information - theoretic security based on laws of quantum mechanics @xcite . here ,
security against forging relies on the impossibility of perfectly distinguishing between non - orthogonal quantum states .
a serious drawback of previous quantum digital signature schemes is that they require long - term quantum memory , making them unfeasible in practice .
we present the first realization of a scheme @xcite that does not need quantum memory and which also uses only standard linear optical components and photodetectors . in our realization , the recipients measure the distributed quantum signature states using a new type of quantum measurement , quantum state elimination @xcite .
this significantly advances quantum digital signatures as a quantum technology with potential for real applications .
digital signatures are used to ensure that messages can not be forged or tampered with .
signed messages are also transferable , meaning that it is unlikely that one recipient accepts a message as genuine , while another recipient , to whom the message is forwarded , rejects it .
this important property is also called non - repudiation ; a sender can not deny having sent a message .
digital signature schemes are different from encryption , which guarantees the privacy of a message .
both are important cryptographic tasks .
quantum key distribution ( qkd ) @xcite can be used to distribute a secret key for information - theoretically secure encryption , and commercial systems are already available @xcite .
analogously , digital signature schemes relying on quantum mechanics @xcite can also be made information - theoretically secure , in contrast to currently used classical digital signature schemes . in this work
we show that quantum digital signature ( qds ) and qkd require similar experimental components and a comparable level of experimental complexity .
protocols for quantum digital signatures have a distribution stage and a messaging stage .
we will describe the case with one sender and two recipients , but this can be extended to more recipients . in the distribution stage ,
the sender , alice , transmits quantum signature states to the recipients , bob and charlie .
she chooses a sequence of @xmath0 states for each possible message that she might later want to send , for a suitable chosen integer @xmath0 , and distributes one copy of each state sequence to each recipient .
the quantum states are randomly chosen from a set of non - orthogonal states , in our realization four coherent states @xmath1 and @xmath2 , with known magnitude @xmath3 .
the chosen phase sequences are analogous to a private key , known only to alice . in the simplest case , to send a one - bit message later on , alice distributes two sequences of states to each of bob and charlie , one corresponding to the possible message 0 " , and one corresponding to the message 1 " . in the subsequent messaging stage , alice accompanies the message she sends with the classical information about the corresponding sequence of quantum states ; in our realization , the sequence of phases . a recipient of a signed message tests that this agrees with the previously distributed quantum signature states , and accepts the message as genuine if there are sufficiently few mismatches for the whole sequence .
similarly , to forward a message , a recipient forwards the message together with the information about the corresponding quantum signature states .
the new recipient again tests for mismatches and verifies if these are few enough .
previous qds schemes @xcite required that recipients store the signature states in long - term quantum memory until the messaging stage .
once a recipient is given the private information about a signature state , say , that it should be equal to some state @xmath4 , the best way to test for a mismatch is to make a quantum measurement with measurement operators @xmath5 .
that is , to test if the state has any component orthogonal to the state it is declared to be .
[ fig : setup ] the requirement for quantum memory is clearly unfeasible at present .
there may be days , weeks , or longer between the distribution and the messaging stages , whereas state - of - the - art quantum memories can not achieve coherence times longer than tens of minutes at room temperature @xcite . a protocol that circumvents quantum memory
was suggested in @xcite , and our current experiment realizes a variant of this scheme .
here , the recipients measure the signature states directly at the end of the distribution stage .
only classical information needs to be stored . in
@xcite , unambiguous state discrimination measurements were envisaged . in our realization , we improve on this idea so that bob and charlie instead use _ unambiguous quantum state elimination
_ @xcite to probabilistically exclude one or more phases for each signature state .
our experimental setup is shown in fig .
[ fig : setup ] .
the state elimination measurement can for coherent states be realized using linear optics and photodetectors .
each recipient uses two detection systems , shown within dashed light blue lines in figure [ fig : setup ] , where the signature states are interfered with reference pulses of phase 0 in the top and phase @xmath6 in the bottom interferometer .
polarization routing @xcite is employed for the orthogonally polarized signal and reference pulses , using the polarizing beam splitter ( pbs ) and combiners ( pbc ) .
the reference pulses enter through the left - hand input ports of beam splitters 2 " and 3 " while the delayed signal pulses enter through the top .
detecting photons in any of the output ports excludes one possible phase , similar to a recent realization of unambiguous state discrimination ( usd ) @xcite . whereas usd requires excluding all but one of the quantum states , we only require elimination of at least one state ( phase ) .
this significantly increases the number of usable signature elements by requiring fewer detection coincidences .
the process of use is summarized here and described in more detail in the appendix [ supp ] . to estimate the resulting advantage ,
assume that the amplitude entering bob s and charlie s measurement setups is @xmath7 , and neglect e.g. phase imperfections .
the probability of excluding the coherent state of opposite phase to the one that is sent is then @xmath8 , and the probabilities to exclude the other two are @xmath9 .
the probability of excluding all three states that were not sent is @xmath10 , while the probability of excluding at least one of them is @xmath11 , which is always greater .
if , as in our experiment , @xmath12 is small , then this quantity is much greater than @xmath10 .
a forger must avoid declaring a phase that has been eliminated ; more precisely , avoid this for sufficiently many signature sequence positions .
if bob ( or charlie ) succeeds in eliminating three of the four possible phases for one signature position , then a forger must select the single remaining phase to avoid a mismatch . but even if just one phase is ruled out , a forger must avoid selecting this phase . with use , therefore , many events which would count as non - detected if using usd will now contribute to the detection of forging .
consequently , using state elimination leads to an improvement in the signature generation rate . for
both use and usd , the forger s probability of avoiding too many mismatches decays exponentially with the signature length @xmath0 .
a more detailed security analysis is found in @xcite and in the appendix [ supp ] .
we have examined security for a single use of the protocol , for general repudiation attacks and all forging attacks except those involving entangling operations on successive signature sequence states .
so - called composable security " remains an important issue . in short , security against forging follows since alice
s signature states are chosen from a set of non - orthogonal quantum states , which can not be distinguished perfectly .
only alice has the full description of these states .
note that the number of recipients depends on protocol parameters , since if too many copies of alice s signature states are available , or if @xmath13 is too large , then the private phases could be determined reliably enough to forge a message unless protocol parameters such as @xmath0 are adjusted .
to prevent repudiation , recipients must ensure that they are sufficiently unlikely to disagree on the validity of a message . here , as in @xcite
, this is achieved using an all - optical fiber multiport , shown in fig .
[ fig : setup ] within dashed brown lines .
bob and charlie split the pulses received from alice using a 50:50 beam splitter .
bob sends to charlie half of the pulse he received from alice , and charlie does correspondingly .
bob then combines the component he received directly from alice with the component he received from charlie on another 50:50 beam splitter , and charlie again does correspondingly .
this symmetrizes bob s and charlie s quantum states for each position in the signature sequences , so that their measurement statistics at the output of the multiport are identical . by choosing a lower allowed fraction of mismatches @xmath14 for accepting a message received directly from alice , and a higher allowed fraction @xmath15 for verifying that a forwarded message is genuine ,
it can be made unlikely that two recipients will disagree on the validity of a message , see @xcite and appendix [ supp ] .
moreover , the multiport guarantees that even if alice uses general , possibly entangled quantum states , she still can not make bob and charlie significantly disagree on the validity of a signature .
in addition , by considering counts at the multiport null - ports , the recipients can guard against certain types of forging . in @xcite and in the appendix [ supp ]
we show that the probabilities for repudiation and forging decay exponentially in @xmath0 , by suitable choice of protocol parameters depending on the properties of an actual implementation .
the scheme can also be made _ robust _ , that is , if all parties are honest , then the protocol runs as intended with high probability . in any implementation
, errors will occur even if all parties are honest . therefore to ensure robustness one should for example select @xmath16 .
defining the level of security in qds is not straightforward , since different parties may be honest and dishonest .
here we assume that one chooses values of @xmath14 and @xmath15 such that the probabilities for repudiation , forging , and for rejection if all participants are honest are all equal ( see appendix [ supp ] ) .
the probability of any of these undesirable events occurring is then @xmath17 where @xmath18 is the gap giving a lower bound on the advantage that someone ( e.g. alice ) has if she knows the signature , compared to someone else ( e.g. a forger ) who makes a guess by performing a measurement on the signature copies ( see appendix [ supp ] ) . in this paper
, we will call the failure probability the _ security level of the qds scheme_. eq .
( [ failure_prob ] ) shows that a greater gap @xmath18 gives better protocol performance .
the figure of merit that we will use to quantify the performance of our experiment is the length of the signature @xmath0 required to sign a `` half - bit '' message for a given security level .
one can also define the _ rate _ of the signature as the number of bits per second that can be signed securely , given the clock rate of the source used .
our experiment uses a clock rate of 100 mhz , due to the temporal response profile of the geiger - mode silicon single - photon avalanche diodes ( si - spads ) @xcite .
we explored amplitudes from @xmath19 to 11 .
coherent states are generated by a temperature - stabilized pulsed vertical cavity surface emitting laser ( vcsel ) with wavelength 850.17 nm , attenuated to the desired mean photon number per pulse @xmath13 , defined at the launch from alice into the multiport .
for a given run of the experiment with some @xmath13 , we registered the phases that bob and charlie ruled out .
this experimental data gives the probabilities of excluding particular states , given that alice sent a certain state .
all losses are included , since these probabilities are determined from the experimentally measured ratio of detection events to the total number of pulses sent by alice . for the qds scheme to be secure , an honest participant must be able to detect a difference between forged and genuine signatures ( see appendix [ supp ] ) . how large this difference is determines @xmath18 ( see @xcite ) , and therefore through eq .
( [ failure_prob ] ) determines the length @xmath0 required for a desired security level .
the gap @xmath18 is proportional to the the transmittance ( one minus the losses ) ( appendix [ supp ] ) and therefore the length @xmath0 for a fixed security level decreases quadratically as the transmittance increases . in short
, the difference between the success and failure probabilities for use determines how well a participant can identify a false declaration .
experimental results are shown in figure [ fig : results ] .
each data point represents the mean of several measurements .
vertical error bars are the standard deviation , and horizontal error bars the uncertainty in the mean photon number due to a pulse - to - pulse variation of @xmath201.5% in vcsel intensity . in figure
[ fig : results]b use success " means that at least one state was eliminated , as long as the state that alice actually sent was not eliminated .
use failure " means that the state that alice actually sent was eliminated .
the success probability for usd is also shown , and is considerably lower than for use . with use ,
one sometimes excludes more than one state .
all usd success " events are included in the use success " data . as already noted , with
use many events which would count as undetected in usd will now contribute to the detection of forging , in addition to all usd success events .
the difference between the success and failure probabilities is greater for use than it is for usd , similarly indicating that use leads to a greater chance of detecting forging .
for all investigated values of @xmath13 , the success probability for use is much higher than the failure rate . for higher failure rates
, one has to set acceptance and verification thresholds @xmath14 and @xmath15 higher to ensure robustness .
this in turn increases the signature length required to ensure the same security level .
the primary cause of a failure " , for both use and usd , was the fringe visibility of the detection setups , which was 80.9% .
the multiport has a fringe visibility of 99.7% . when determining the optimal @xmath13 , one has to consider that the gap also depends on @xmath21 , which is the minimum error probability that a forger obtains if he tries to guess alice s signature by measuring a copy of the quantum signature ( appendix [ supp ] ) .
for very small @xmath3 , @xmath21 is large , but detecting a false declaration is difficult , while for very large @xmath3 , @xmath21 is small but detecting a false declaration is relatively easy . since the ability to detect a false declaration
is estimated from experimental data , and does not have an analytical expression , it is not straightforward to determine the optimum @xmath3 . in our experiment
, the best gap @xmath22 occurs for @xmath19 . for a security level of @xmath23
this gives @xmath24 to sign a `` half - bit '' .
this is an impractical signature length , and below we will comment on planned improvements in order to make this rate more practical .
the signature length @xmath0 increases with increased distance between parties , since @xmath18 in eq .
( 1 ) is proportional to the transmittance @xmath25 .
for example , if @xmath25 is squared , then the @xmath0 required for the same level of security will increase by a factor of @xmath26 . in any event , if honest recipients see a difference between a forged and a genuine signature , however small , then it is always possible to find values of @xmath27 and @xmath0 to give a desired level of security . to conclude , we have experimentally demonstrated a first realization of a qds scheme which does not require long - term quantum memory , and where the recipients use quantum state elimination .
this is an important step in developing practical qds systems .
our experiment uses phase - encoded coherent states .
recently , arrazola and ltkenhaus suggested using phase - encoded coherent states for quantum fingerprinting @xcite . in our demonstration , due to the difficulty of stabilizing a multiport with long optical paths
the sender and receivers were only separated from each other by approximately five meters of optical fiber .
separate reference signals are needed for calibration before signature transmission , and as phase reference for the use measurements .
tampering with reference pulses by a malevolent party should not lead to higher probability of forging or repudiation than tampering with signal states themselves ( appendix [ supp ] ) .
also , reference signals can be bright , and thus can in principle be fully monitored through quantum tomography .
we are currently exploring three changes to significantly improve performance .
first , by extrapolating data from a recent experiment on use , we expect the optimal @xmath13 to be around @xmath28 . due to the high losses of this early prototype we were unable to successfully resolve measurements at this @xmath13 .
the second improvement is to use a protocol that does not require a multiport , to decrease loss .
non - repudiation then needs to be guaranteed in an alternative way , similar to our recently proposed alternative qds schemes @xcite , which could be modified to use phase - encoded coherent states , similar to the current realization .
we estimate that implementing these changes will result in a gap of @xmath29 , and length @xmath30 for a security level of @xmath23 .
this protocol also potentially offers increased distances between sender and receivers .
finally , increasing the clock rate , and therefore the transmission rate , is possible .
the phase modulators , vcsel , and driving electronics are capable of clock - rates up to 3.3 ghz . in the system described in this letter
we did not employ such clock - rates due to the limitations of the time - stamping electronics @xcite .
this work was supported by the uk engineering and physical sciences research council ( epsrc ) through grants ep / g009821/1 , ep / k022717/1 , and the epsrc doctoral prize fellowship grant . p.w . gratefully acknowledges partial support from cost action mp1006 . |
the robert c. byrd green bank telescope ( gbt ) is the premiere single - dish radio telescope operating at meter , centimeter and millimeter wavelengths [ 1 ] . with an off - axis parabolic reflector of size 100 m
@xmath0 110 m , the gbt is the largest fully - steerable radio telescope in the world .
its unblocked aperture gives it the sensitivity of a much larger antenna , flat spectral baselines and higher sensitivity to low surface brightness emission .
the gbt is located at the national radio astronomy observatory ( nrao ) site in green bank , west virginia , usa .
it is being used for a variety of unique experiments over a broad range of science . since its commissioning in 2000
, the gbt s capabilities have been greatly expanded in frequency coverage as well as the instantaneous field of view for observations .
the field of view is enhanced using focal plane arrays .
the new 4 mm receiver is an example of expansion in frequency coverage of the gbt .
so far observers have been using all these receivers with an existing spectrometer backend , which has only 3 level sampling , 800 mhz bandwidth and a minimum integration time of one sec .
the bandwidth supported by this spectrometer for focal plane arrays is @xmath1 50 mhz .
thus the capabilities of the existing spectrometer do not match with those of the new receivers .
it is now time to upgrade the backend capability of the gbt . the national science foundation advanced technologies and instrumentation ( nsf - ati ) program has funded a new spectrometer backend for the gbt .
this spectrometer is being built by the cicada collaboration a collaboration between the nrao and the center for astronomy signal processing and electronics research ( casper ) at the university of california , berkeley .
the new spectrometer vastly enhance the capability of the gbt to support several unique science projects .
the projects include mapping temperature and density structure of molecular clouds ; searches for organic molecules in the interstellar medium ; determination of the fundamental constants of our evolving universe ; redshifted spectral features from galaxies across cosmic time and survey for pulsars in the extreme gravitational environment of the galactic center .
one of the design considerations for the new spectrometer is to process larger bandwidth ( 1.25 ghz ) with the focal plane array ( fpa ) receivers .
the k - band ( 18 26.5 ghz ) fpa of the gbt has 7 feeds .
thus the spectrometer is designed to process 8 ( 7 + 1 spare ) dual polarized signals .
these signals are processed by 8 independent spectrometer units working in parallel .
a block diagram of the spectrometer is shown in fig . [ fig1 ] .
the 8 spectrometer units are connected to the converter racks ( cr ) outputs , which is the last stage of the gbt s intermediate frequency ( if ) system .
the spectrometer can broadly be divided into two parts : ( 1 ) fpga based hardware and ( 2 ) pipeline computing facility .
casper s roach board is used as the fpga based hardware .
the pipeline computing will be implemented using a cluster computing facility ( hpc ) .
the analog signals from the cr , after amplification and band limiting to 1.5 ghz , are fed to 8-bit adcs attached to the roach boards .
the sampling clock for the adc is generated using a synthesizer , which is locked to the observatory 10 mhz standard .
the fpgas in the roach board process the digitized signal and send spectra and cross correlations of the two polarizations to the hpc through a 10 gbe switch .
the hpc consists of 8 ( + 1 spare ) independent pcs .
each pc will have a dual port 10 gbe card and a gpu gtx480 card .
the gpu will be used for implementing observing modes with bandwidth @xmath1 250 mhz and the 8 sub - band capability ( see section [ mode ] ) .
one of the ports of the 10gbe card will receive data from the roach boards and the second port will send data to the data storage system .
the hpc is connected to the data storage system through the same 10 gbe switch to which the roach boards are connected .
the data storage system will be a parallel file system ( eg .
lustre ) based storage and will support @xmath2 100 tb of disk space .
for all the specified observing modes data will be written to this storage system .
the gbt users will be accessing data from this storage system through a 1 gbe link as shown in fig [ fig1 ] . the monitor and control ( m&c ) of the whole spectrometer
is done through a dedicated computer .
the m&c system will obtain data from the roach board through a 1 gbe ( or 100 mbe ) network .
a 10 gbe link to this network provides the m&c information from the hpc .
a 1 pps signal from the observatory time standard is connected to the roach boards .
this 1 pps is the main synchronization signal for the spectrometer .
switching signals are used during observations to synchronize spectral measurements with noise and/or frequency switching .
the internal switching signals are generated by one of the roach boards ( switching master ) and will be sent to other roach boards through a distribution system .
the switching master will also accepts external switching signals and will send them to other boards through the same distribution system . the spectrometer software architecture
is shown in fig .
the software components can be broadly divided into two parts the m&c and the observing pipeline .
the m&c is designed based on the existing gbt software system . in this system
each hardware device is controlled by a ` manager ' .
the manager also sends the monitoring information at regular intervals .
they also have the functionality to record data to a fits file . in the spectrometer design , the roach
m&c managers control the roach boards and the hpc m&c managers control the data acquisition and flow in the hpc units .
the spectra from the roach are averaged to the required integration time and put in a shared memory by the data acquisition module .
the hpc m&c manager reads the data from the shared memory and records to the disk storage system . for fpa observations , the hpc m&c manager sends data to the data processing pipeline .
this pipeline will produce a uniformly gridded , calibrated spectral map of the observing region as its end product . the operations of this pipeline can be divided into beam based calibration and mapping process . the beam based calibration process ( gbt calibration pipeline ) will be implemented in the hpc . after calibration , the hpc sends the data to a computer ( arcturus ) where the spectral image is made by the mapping process .
the spectrometer supports a variety of observing modes .
the major modes of operation are single and 8 sub - band modes . in single sub - band mode the maximum bandwidth a spectrometer unit can process is 1.25 ghz .
there will be 1024 spectral channels across this bandwidth which gives a spectral resolution of @xmath2 1.5 khz .
the minimum integration time supported by this mode is 0.5 msec .
the minimum bandwidth provided in single sub - band mode is 1 mhz with spectral resolution of 30 hz ( ie 32768 channels ) .
the minimum integration time for this mode is about 10 msec .
the spectrometer will support 8 fully tunable sub - bands within the 1.25 ghz bandwidth ( see fig .
[ fig3]b ) .
the center frequency of the sub - band is tunable to an accuracy of 10 khz .
the sub - band bandwidth varies from 30 mhz to 1 mhz .
the number of spectral channels per sub - band for these different bandwidths is 4096 .
the 8 spectrometer units can be configured in different observing modes . they can also be used together to process a contiguous bandwidth of up to 10 ghz .
an example of using the spectrometer in this configuration is shown in fig .
the total number of sub - bands available when all the 8 units are used together is 64 . for fpa observations , all the 8 spectrometer units
will usually be configured to the same observing mode ( see fig . [
we thank chris clark , eric sessoms , jay lockman , mark clark , richard lacasse , roger norrod , ron maddalena , steven white and wolfgang baudler for their informative comments and suggestions at various stages of the spectrometer project .
prestage , r. m. , et . al.,the green bank telescope , " _ proceedings of the ieee _ ,
vol 97 , issue 8 , 2009 , pp .
1382 - 1390 + | a new spectrometer for the green bank telescope ( gbt ) is being built jointly by the nrao and the casper , university of california , berkeley .
the spectrometer uses 8 bit adcs and will be capable of processing up to 1.25 ghz bandwidth from 8 dual polarized beams .
this mode will be used to process data from focal plane arrays .
the spectrometer supports observing mode with 8 tunable digital sub - bands within the 1.25 ghz bandwidth .
the spectrometer can also be configured to process a bandwidth of up to 10 ghz with 64 tunable sub - bands from a dual polarized beam .
the vastly enhanced backend capabilities will support several new science projects with the gbt . |
a galactic plane survey was performed in 2004 by the h.e.s.s . imaging atmospheric cerenkov telescope ( iact ) with a flux sensitivity of 0.02 crab for gamma rays above 200gev @xcite .
fourteen new gamma - ray sources were detected at significance levels above 4@xmath8 , and 11 of the sources either have no counterpart or possible counterparts with significant positional offsets .
hess j1804@xmath0216 is one of the brightest , and its spectrum is softest in this survey ; the flux is about 0.25 crab above 200gev with a photon index of @xmath9 .
in addition , with a size of @xmath10 arcmin , it is one of the most extended tev gamma - ray sources .
the h.e.s.s .
collaboration proposed two possible counterparts : the supernova remnant ( snr ) [email protected] and the young vela - like pulsar b1800@xmath021 .
however , the tev gamma - ray source does not coincide exactly with either of these .
[email protected] appears as a larger circular region with a diameter of @xmath11 arcmin , and the geometric center @xcite has a large offset , @xmath12 arcmin , from the centroid of hess j1804@xmath0216 to the northeast , while psr b1800@xmath021 has an @xmath12 arcmin offset to the west .
after the detection of hess j1804@xmath0216 , an snr , [email protected] , was discovered at radio wavelengths @xcite to be located within the error circle of hess j1804@xmath0216 , but with a smaller size , @xmath13 .
_ suzaku _ deep observations discovered two new x - ray sources , suzaku j1804@xmath02142 ( hereafter src1 ) and suzaku j1804@xmath02140 ( hereafter src2 ) , that are near the center of hess j1804@xmath0216 @xcite .
_ chandra _ also detected these sources @xcite .
_ swift _ found three faint x - ray sources in the region of hess j1804@xmath0216 @xcite .
one of them is positionally coincident with a bright star , and another could also be associated with a star close to the boundary of the xrt error circle .
the other positionally coincides with suzaku src2 .
so , there remain 5 possible counterparts : snr [email protected] , psr b1800@xmath021 , snr [email protected] , suzaku src1 , and suzaku src2 .
we briefly describe these sources in the following paragraphs .
snr [email protected] : [email protected] is associated with the w30 complex , which comprises extended radio emission with a number of superposed smaller discrete emission regions @xcite .
radio recombination - line observations have been used to identify discrete sources as h@xmath14 regions , and co observations also show molecular gas to be associated with w30 @xcite .
@xcite reported that massive star formation may be occurring in molecular clouds in w30 . @xcite and @xcite clearly established that [email protected] was an snr by the detection of a nonthermal extended radio emission .
a _ rosat _
observation revealed diffuse x - ray emission only from the northern half of the remnant in the 0.12.4kev band @xcite .
the distance to [email protected] was estimated using several methods .
based on kinematical distances to the h@xmath14 regions associated with the snr , the distance was estimated to be @xmath15kpc @xcite .
@xcite pointed out that more recent galactic rotation models applied to the h@xmath14 regions suggest a near kinematical distance of about 4.8kpc .
they also estimated the distance based on a sedov solution @xcite from the observed x - ray temperature and the angular radius to derive [email protected] for an assumed initial energy of @xmath17 ergs . in this paper
, we adopt @xmath18kpc .
they also estimated the age of the snr , based on a sedov solution from the x - ray observation , to be 1.52.8@xmath19 years under the assumption of an initial energy of @xmath17 ergs , and 2.73.9@xmath19 years under the assumption of a distance of 6kpc .
similarly , we estimated an age of 2.23.1@xmath19 years under the assumption of a distance of 4.8kpc .
@xcite represented an age of @xmath20 years from the relation between the age and the surface brightness in the radio band .
psr b1800@xmath021 : the young vela - like pulsar b1800@xmath021 was found in a radio observation @xcite .
the offset from the centroid of hess j1804@xmath0216 is @xmath2111 arcmin .
the spin period , the spin period derivative , and the characteristic age are @xmath22ms , @xmath23s / s , and @xmath24kyears , respectively @xcite .
the resulting spin - down luminosity is @xmath25 ergss@xmath5 , where @xmath26 is the moment of inertia in units of g@xmath27 .
the distance was estimated to be 3.9kpc from the pulsar s dispersion measure of @xmath28pc@xmath29 @xcite .
the newer cordes & lazio ne2001 model @xcite gives a similar distance of 3.84@xmath30kpc .
we adopt @xmath31kpc throughout this paper .
the association between psr b1800@xmath021 and [email protected] has been discussed in several papers ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , a recent proper - motion measurement @xcite has shown that the pulsar was born outside the currently observed snr , and that is moving more nearly toward the center of [email protected] , rather than away from it , which makes their association very unlikely .
based on a 10ks observation with the _ rosat _ pspc , finley & gelman ( 1994 ) reported a faint x - ray source near the radio pulsar position , and attributed this emission to psr b1800@xmath021 .
recently , @xcite and @xcite reported that an x - ray nebula around the pulsar was detected with _
chandra_. additionally , @xcite reported that the x - ray nebula has two structures : a brighter compact ( @xmath32 ) component ( the inner pulsar wind nebula ( pwn ) ) and an extended ( @xmath33 ) fainter emission component ( the outer pwn ) .
these are asymmetric to the pulsar position and extended toward hess j1804@xmath0216 .
snr [email protected] : snr [email protected] was found in a 90 cm multi - configuration very large array survey of the galactic plane .
the size is @xmath13 , and the offset from the centroid of hess j1804@xmath0216 is 7 arcmin .
the morphology is shell - like and the spectral index is @xmath34 for @xmath35 ( brogan et al .
2006 ) .
suzaku src1 : the offset from the centroid of hess j1804@xmath0216 is 3 arcmin .
@xcite reported that src1 is point - like or compact compared to the spatial resolution of _ suzaku _ with a half - power diameter of about 2 arcmin .
however , kargaltsev et al . ( 2007b ) reported that src1 is extended or multiple ( @xmath36 ) with a _
the _ suzaku _
spectrum was fitted with an absorbed power - law model @xcite .
the best - fit absorbing column is consistent with the galactic hydrogen column in that direction .
since the photon index of @xmath37 is very flat , @xcite suggest that this source is likely to be a high - mass x - ray binary ( hmxb ) . in the _ chandra _ observation @xcite ,
no spectral fitting was able to be performed because of the low signal - to - noise ratio .
the unabsorbed flux , which was estimated from the _ chandra _ observation using the best - fit parameters reported by @xcite , is a factor of @xmath38 1.7 smaller than that reported with the _ suzaku _ observation of @xcite .
the difference could be due to unaccounted systematic errors , or the variability of the source , which supports the hmxb interpretation @xcite .
suzaku src2 : the offset from the centroid of hess j1804@xmath0216 is 1.8 arcmin . though @xcite reported that src2 is extended or multiple , based on the _ suzaku _ observation , @xcite reported that src2 is point - like in a _
chandra _ observation .
this could mean that the more sensitive _ suzaku _ observations have detected a fainter extended pwn component .
the _ suzaku _
spectrum was fitted with an absorbed power - law model @xcite , and the best - fit absorbing column is about an order - of - magnitude higher than the expected galactic column .
this implies that src2 is embedded in dense gas .
spectrum was also well - fitted with an absorbed power - law model , and the obtained absorbing column density is a factor of 23 larger than the galactic column @xcite .
@xcite stated that the large absorption suggests that src2 is located within ( or even beyond ) the galactic bulge , or it shows an intrinsic absorption , which is often seen in the x - ray spectra of hmxbs .
@xcite also reported a marginal pulsation of 106 s in src2 , which supports an hmxb interpretation . on the other hand
, @xcite suggested that src2 is a pwn or a shell - like snr because of the extended morphology observed with _
suzaku _ and the best - fit photon index of 1.7 ( 0.73.1 ) .
none of the above five sources morphologically match hess j1804@xmath0216 .
the counterpart is therefore still unknown . in this paper , we present tev gamma - ray observations of hess j1804@xmath0216 with the cangaroo - iii telescopes and discuss the radiation mechanism and the counterpart by combining our result with multi - wavelength observations .
cangaroo - iii is an array of four iacts , located near woomera , south australia ( 136@xmath39e , @xmath40s , 160 m a.s.l . ) .
each telescope has a 10 m diameter reflector made up of 114 segmented frp spherical mirrors mounted on a parabolic frame @xcite .
the telescopes are situated at the corners of a diamond with sides of @xmath21100 m @xcite .
the oldest telescope , t1 , which was the cangaroo - ii telescope , was not used due to its smaller fov and higher energy threshold .
the imaging camera systems on the other three telescopes ( t2 , t3 and t4 ) are identical , with 427 pmts and a fov of 4.0@xmath41 @xcite .
the pmt signals were recorded by charge adcs and multi - hit tdcs @xcite .
the observations of hess j1804@xmath0216 were made from may to july , 2006 , using the ` wobble ' mode in which the pointing position of each telescope was shifted in declination by @xmath420.5@xmath41 from the centroid of hess j1804@xmath0216 .
the mean zenith angle of the observation was 21@xmath41 , and the total observation time was 86.8 hr . we used the 3-fold coincidence data taken at zenith angles of less than 40@xmath41 . to trigger data recording , an individual telescope was required to have more than four pixels receiving over 7.6 photoelectrons within 100ns ( local trigger ) , with a global trigger system to determine the coincidence of any two of the three telescopes @xcite .
we rejected data taken under bad weather conditions in which the shower event rate was less than 6hz .
taking into account the daq dead - time , the effective live time was calculated to be 76 hr .
the basic analysis procedures are described in detail in @xcite and @xcite . using calibration data taken daily with leds , the recorded charges of each pixel in the camera were converted to the number of photoelectrons . at this step
we found 7 bad pixels out of 427 pixels for t2 , 5 for t3 , and 1 for t4 , due to their higher or lower adc conversion factors in these observations .
these bad pixels were removed from this analysis , which was also reflected in the monte carlo simulations .
after that , every shower image was cleaned through the following cangaroo - iii standard criteria .
only pixels that received @xmath435.0 photoelectrons were used as `` hit pixels '' .
then , five or more adjacent hit pixels , with arrival times of within 30ns from the average hit time of all pixels , were recognized as a shower cluster . before calculating image moments
the hillas parameters " @xcite we applied the `` edge cut '' described in @xcite .
we rejected events with any hits in the outer - most layer of the camera .
the orientation angles were determined by minimizing the sum of the squared widths with a constraint given by the distance predicted by monte carlo simulations .
we then applied the fisher discriminant method @xcite with a multi - parameter set of @xmath44 , where @xmath45 and @xmath46 are the energy corrected width and length , and the suffixes represent the telescope ids .
the fisher discriminant ( fd ) is defined as @xmath47 , where @xmath48 is a set of coefficients mathematically determined in order to maximize the separation between two fds for gamma rays and hadrons . for a background study we selected a ring region around the target , @xmath49deg@xmath50 , where @xmath51 is the angular distance to the centroid of hess j1804@xmath0216 reported by the h.e.s.s .
group @xcite , and obtained the fd distributions for the background , @xmath52 , and monte carlo gamma rays , @xmath53 .
finally , we could fit the fd distributions of the events from the target with a liner combination of these two components .
the observed fd distributions , @xmath54 , should be represented as @xmath55 , where @xmath6 is the ratio of gamma - ray events to the total number of events . here ,
only @xmath6 was optimized .
this analysis method was verified by an analysis of the crab nebula data taken in december , 2005 .
the reflectivities of each telescope , which were used in the monte carlo simulations , were monitored every month by a muon ring analysis of a calibration run taken individually by each telescope .
we obtained relative light - collecting efficiencies with respect to the original mirror production times of 0.60 , 0.60 and 0.65 for t2 , t3 and t4 , respectively . throughout this analysis
, we used the monte carlo simulations for gamma rays assuming a point - source .
the obtained @xmath56 plot is shown in fig.[theta ] with the point spread function ( psf ) of our telescopes , @xmath57 ( 68% containment radius ) .
the numbers of excess events that we detected above 600gev were @xmath58 within @xmath59 deg@xmath50 , based on the assumption that it was a point source , @xmath60 within @xmath61 deg@xmath50 , which corresponds to that used in the spectral analysis by h.e.s.s .
( and taking into account the difference between our psf and that of h.e.s.s . ) , and @xmath62 within @xmath63 deg@xmath50 . the tev gamma - ray emission is extended , and the morphology of gamma - ray like events , derived with boxcar smoothing with each pixel replaced by the average of its square neighborhood , is shown in fig.[map ] .
the number of excess events was individually estimated by the fd - fitting method in each @xmath64 sky bin .
when we evaluated the outer regions ( @xmath65 deg@xmath50 ) , we had to consider gradual deformations of the fd distributions at larger angular distances from the target .
therefore , we selected an annulus with radii @xmath66 centered on the evaluated region as the background .
for the inner regions , @xmath67 deg@xmath50 , the events in @xmath68deg@xmath50 , excluding the evaluated region , were adopted as a background .
the intrinsic extent of the tev gamma - ray emission was estimated by a 2d gaussian fit on our unsmoothed excess map .
the intrinsic deviations along the right ascension and declination axes were calculated to be @xmath69 and @xmath70 , respectively .
the best - fit centroid position was obtained ( r.a , dec [ j2000])@xmath71(271.079@xmath72 , -21.727@xmath72 ) .
the offset from the best - fit position reported by h.e.s.s .
@xcite is ( @xmath73r.a , @xmath73dec)@xmath71(@xmath74 , @xmath75 ) .
the offset is not significant given our psf .
figure [ flux ] represents a reconstructed vhe gamma - ray differential spectrum compatible with a single power - law : @xmath2(e/1 tev)@xmath3 @xmath4s@xmath5tev@xmath5 with a photon index @xmath6 of @xmath7 . to obtain the spectrum
, we used a cut of @xmath76 deg@xmath50 .
the relevant systematic errors are due to the atmospheric transparency , night sky background fluctuations , uniformity of camera pixels , and light - collecting efficiencies .
in addition , the signal integrating region was changed from @xmath77 deg@xmath50 to 0.3 deg@xmath50 , and the difference in fluxes was incorporated in the systematic errors . the tev gamma - ray extension and
the flux obtained by cangaroo - iii were consistent with those by h.e.s.s .
our result indicates that the tev gamma - ray emission was unchanged between the h.e.s.s.observations in 2004 and ours in 2006 .
as described in 1 , there are 5 possible counterparts : snr [email protected] , psr b1800@xmath021 , snr [email protected] , suzaku src1 , and suzaku src2
. figures [ sed_allplot ] and [ sed_xray ] show the spectral energy distribution ( sed ) of these counterparts and hess j1804@xmath0216 including the results of this work .
we extend the introductions to these sources given in 1 with a more detailed summary of the sources and their characteristics in the following paragraphs .
@xcite state that suzaku src1 and src2 are physically associated with hess j1804@xmath0216 .
they suggested that x - rays and tev gamma - ray emission could come from an snr shock , based on a model proposed by @xcite ; in an old snr with an age of @xmath78 years , primary electrons have already lost most of their energy , and only nucleonic cosmic rays remain .
additionally , the old snr shock colliding with a giant molecular cloud ( gmc ) can emit hard nonthermal x - rays from secondary electrons and strong tev gamma rays from shock accelerated protons through @xmath79 decay .
this model can explain the observed large ratio of the tev gamma - ray to x - ray flux ( factors of @xmath21100 ) .
the large x - ray absorption of src2 also supports this scenario .
@xcite suggested that a pwn is the source of hess j1804@xmath0216 , like hess j1825@xmath0137 or vela x. hess j1825@xmath0137 is likely to be associated with the pwn [email protected] around the vela - like pulsar b1823@xmath013 . the tev gamma - ray emission detected with h.e.s.s .
covers a much larger area than the x - ray emission from [email protected] , extending up to 1@xmath72 southward from the pulsar @xcite .
however , both the tev gamma - ray and the low surface - brightness x - ray emission have similarly asymmetric shapes , and they are offset in the same direction with respect to the pulsar position .
a similar picture is observed around the vela pulsar @xcite .
these phenomena can be explained by the `` crushed pwn '' hypothesis @xcite : on a time scale of @xmath80 years , the reverse snr shock front propagates toward the center of the remnant , where it crushes the pwn , and asymmetries in the surrounding interstellar medium give rise to an asymmetric shape and offset of the pwn relative to the pulsar and explosion site .
@xcite considered the possibility that suzaku src1 or src2 are a pwn powering hess j1804@xmath0216 .
however , the 3.24 s time resolution of the _ chandra _ acis observation precludes a search for the subsecond pulse periods expected for a young pulsar .
therefore , there is no strong evidence to support it at this point .
they also suggested that psr b1800@xmath021 is associated with hess j1804@xmath0216 .
its asymmetric pwn component extending toward hess j1804@xmath0216 , detected by _
@xcite , shows a hint of the association , but the sensitivity of the _ chandra _ observation was possibly insufficient to detect the pwn beyond 15@xmath8120@xmath81 from the pulsar .
additionally , they pointed out that the extended morphology of hess j1804@xmath0216 argues against the hmxb interpretation because of the weak observational evidence for hmxbs producing extended tev gamma - ray emission .
@xcite concluded that psr b1800@xmath021 can not account for the spectrum of hess j1804@xmath0216 , and [email protected] is probably the source of the tev gamma rays .
however , they considered only a pion - decay model without any consideration of the inverse compton process , and for psr b1800@xmath021 they only considered the acceleration of charged particles across voltage drops in the relativistic winds near the light cylinder .
additionally , they did not take into account the possibility of other sources besides psr b1800@xmath021 and [email protected] .
based on the above discussions and the sed , we now discuss the radiation mechanism of hess j1804@xmath0216 and its counterpart .
for suzaku src1 , @xcite reported a very hard photon index of @xmath82 , which corresponds to the power - law index of electrons of @xmath83 .
this unusual value indicates that the x - ray emission is not synchrotron radiation .
therefore , we do not discuss the sed of src1 .
@xcite suggested that src1 might be an hmxb .
if it were , the association between src1 and hess j1804@xmath0216 would be unlikely as discussed in @xcite .
the results of independent observations of the h.e.s.s.and cangaroo - iii telescopes show that hess j1804@xmath0216 is quite extended ( @xmath84 ) .
this precludes the possibility that hess j1804@xmath0216 is an active galactic nucleus .
the plausible candidates seem to be an snr or a pwn .
the snr candidates associated with hess j1804@xmath0216 are snr [email protected] , snr [email protected] , and suzaku src2 .
* snr [email protected] : * we discuss the association between hess j1804@xmath0216 and snr [email protected] .
figure [ sed_diffuse ] shows the leptonic model curves used to fit the tev gamma - ray spectra obtained by h.e.s.s . and cangaroo - iii .
we examined plausible inverse compton ( ic ) model curves on the assumption that the electron spectrum is a single power - law with an exponential cutoff , @xmath85 where @xmath86 is the normalization factor , @xmath87 is the electron energy , @xmath88 is the spectral index of the injected electrons , and @xmath89 is the maximum electron energy . in this paper
, we do not consider hard spectra of electrons with @xmath90 . to obtain the entire ic model curve , we used ir and optical ( starlight ) photon fields for the target photons in addition to the cosmic microwave background ( cmb ) field density . here , we used an interstellar radiation field ( isrf ) derived from the latest ( v50p ) galprop package @xcite .
the isrf was given for three components ( cmb , ir from dust , and optical starlight ) as a function of the distance from the galactic center , @xmath91 ( in kpc ) , and the distance from the galactic plane , @xmath92 ( in kpc ) .
we extracted the spectra at @xmath93 at the snr [email protected] position , as shown in fig .
the x - ray upper limits for diffuse emission constrained @xmath88 to be less than 2.5 for the ic model curves .
we obtained @xmath89 to fit the ic model curves to the tev gamma - ray spectra for each fixed spectral index @xmath88 ( table [ sum_ic ] ) .
the electrons causing the ic scattering also emit synchrotron radiation . in the radio band ,
the spectrum of the whole region of [email protected] is regarded as the upper limits . here , the upper limits in the near - infrared ( nir ) band , in fig .
[ sed_allplot ] , are neglected because they are for point sources .
the magnetic field is constrained by the radio upper limits and the x - ray upper limits for diffuse emission .
table [ sum_ic ] lists the upper limits of the magnetic field , maximum electron energy , and the total energy of electrons above 0.51 mev , @xmath94 at @xmath95kpc for each electron spectral index @xmath88 .
@xmath96 is estimated to be less than @xmath97 ergs for @xmath98 .
although the total explosion energy of some supernovae are much higher , for example , the total energy of sn2003lw , a type ic supernova , is estimated to be @xmath99ergs @xcite , the typical total energy of a supernova explosion is estimated to be @xmath100ergs .
if the efficiency of the energy to accelerate the electrons is 10% , the obtained @xmath96 can satisfy the snr origin scenario .
additionally , the hadronic scenario can also explain the tev gamma - ray spectrum for [email protected] . in fig .
[ sed_pion ] , we show the @xmath79 decay model curves with the assumption that the proton spectrum is a single power - law with an exponential cutoff , @xmath101 where @xmath102 is the normalization factor , @xmath103 is the proton energy , @xmath104 is the spectral index of the injected protons , and @xmath105 is the maximum proton energy . in this paper , we do not consider hard proton spectra with @xmath106 .
the egret upper limit @xcite constrained @xmath104 to be less than 2.3 .
we obtained @xmath105 to fit the @xmath79 decay model curves to the tev gamma - ray spectra for each fixed spectral index @xmath104 .
table [ summary ] lists the fitting parameters and the total energy of protons above 1gev , @xmath107 at a distance for snr [email protected] of 4.8 kpc with the assumption that the interstellar medium ( ism ) density is @xmath108 @xmath29 .
the total energy of protons , @xmath109 , is @xmath110 kpc)@xmath111 @xmath112ergs with @xmath113 . for a typical total supernova explosion energy of @xmath100ergs , if the efficiency of the energy to accelerate the protons is 10% , the ism density should be @xmath114@xmath29 . with the assumption of @xmath115 , @xmath109 is @xmath116 kpc)@xmath111 @xmath112 ergs , and the ism density should be @xmath117 @xmath29 for an efficiency of 10% .
however , @xcite derived the electron density @xmath118 to be @xmath119 kpc)@xmath120 @xmath29 for an x - ray emitting gas in the remnant , based on the _ rosat _ observation .
in addition , the lack of the x - ray emission in the tev gamma - ray emission region , shown in fig . [ map ] , indicates that the ambient density at the region is lower than that at the x - ray emitting region , that is , @xmath121@xmath29 . if it were , the hadronic scenario would be unlikely . on the other hand ,
a gmc , srby3 , is near ( in projection ) to hess j1804@xmath0216 .
the offset , the angular radius , the distance , and the mass of the cloud are @xmath122 , @xmath123 , 5.3 kpc , @xmath124 , respectively @xcite .
the mean number density is calculated to be @xmath125@xmath29 under the assumption that it is spherical symmetry .
if it is associated with hess j1804@xmath0216 , there can be a high - density medium , and the hadronic scenario can be satisfied . additionally , we estimated @xmath126 , the number ratio of protons to primary electrons . under the assumption that the spectral indices of the protons and primary electrons are the same , @xmath127 , @xmath126
is derived from eqs.([singlepl_el])([singlepl_pro ] ) as @xmath128 .
generally the maximum electron energy @xmath89 is the same as @xmath105 , or @xmath89 is lower than @xmath105 due to a cooling effect , that is , @xmath129 . in fig .
[ sed_pion ] , we show the leptonic model curves with the assumption of @xmath130 @xmath112 , b = 3 @xmath131 g , @xmath132 , and @xmath133 tev .
the leptonic model curves move up with the lower @xmath126 . with the assumption of b = 3@xmath131 g
, @xmath126 is constrained to be @xmath134 @xmath112 , because the ic model curve can not exceed the tev gamma - ray spectrum .
if a gmc is associated with hess j1804@xmath0216 , and @xmath114 @xmath29 , the obtained @xmath126 is consistent with that of the average cosmic rays in our galaxy , @xmath135 . on the other hand , for a high @xmath136 of @xmath137
, we need to consider the contribution of the emission from secondary electrons produced by charged pions @xcite . however , a detailed discussion about secondary electrons is beyond the scope of this paper .
* suzaku src2 : * for suzaku src2 , we used the overlapping error region obtained by _ suzaku _ and _ chandra_. in fig .
[ sed_point ] , we examined plausible synchrotron model curve within the error of the unabsorbed x - ray flux on the same assumption as eq .
( [ singlepl_el ] ) with adopting @xmath138=3@xmath131 g and @xmath132 .
the obtained ic model curves for the same electron spectrum can not reproduce the tev gamma - ray spectrum .
even if the higher ir energy density than that in galprop is adopted , the cutoff energy of the models does not change from @xmath139 tev , while the observed cutoff energy should be @xmath140 tev ; such a low cutoff energy is due to the soft spectrum of hess j1804@xmath0216 .
a stronger magnetic field can reduce the cutoff energy of the ic model curves . in order to make the cutoff energy of 0.3tev for suzaku src2 ,
the magnetic field should be more than 7 mg , and to fit the tev gamma - ray spectrum with the magnetic field the ir energy density should be @xmath141 times higher than that in galprop .
these values are unlikely .
we therefore conclude that the tev gamma - ray spectra are not produced by the ic model curves for the electron spectrum .
we also examined the bremsstrahlung emission for the same electron spectrum . to fit the tev gamma - ray spectrum ,
the ism density should be more than @xmath142@xmath29 , and the magnetic field should be stronger than 1.8 mg for suzaku src2 .
due to such incredibly high values of the ism density and the magnetic field , we also reject the scenario that the bremsstrahlung radiation produces the tev gamma - ray spectrum . a simple solution to explain
the observed tev gamma - ray spectrum is that accelerated protons produce the gamma rays , as shown in fig .
[ sed_point ] .
therefore , the tev gamma rays and the x - rays could be produced by accelerated protons and electrons , respectively .
assuming the spectral index , @xmath143 , the maximum electron energy is @xmath144tev ( table [ summary ] ) , and this maximum energy makes the lower limit of the magnetic field of @xmath145@xmath131 g to fit the x - ray data . with the magnetic field , we calculated the lower limit of @xmath146 @xmath112 .
this value is much higher than that of the average cosmic rays in our galaxy , @xmath135 .
* [email protected] : * we consider the possibility that hess j1804@xmath0216 is associated with [email protected] .
for the leptonic scenario , it is difficult to explain the fact that the size of [email protected] in the radio band is much smaller than that in the tev emission region . on the other hand ,
if the accelerated protons produced the tev gamma rays , the difference in the size could be explained due to the difference of the diffusion length between protons and primary electrons .
however , we need additional information , either on the distance or from other wavelengths , to discuss the energetics or @xmath126 . at the present time , pwne are the largest population of identified tev galactic sources , and the number of tev gamma - ray sources that are located near a plausible pulsar candidate are increasing @xcite .
therefore , it is very important to consider the pwn origin .
the pwn candidates associated with hess j1804@xmath0216 are the pwn of psr b1800@xmath021 , and suzaku src2 .
* pwn of psr b1800@xmath021 : * to estimate the total energy supplied from a pulsar , we should consider any braking effects for spin - down luminosity . if a pulsar spins down from an initial spin period of @xmath147 according to @xmath148 assuming both @xmath149 and @xmath150 are constant , where @xmath149 is the braking index of the pulsar , the spin - down luminosity , @xmath151 , is given as @xmath152 @xmath153 where @xmath154 is the initial spin - down timescale @xcite .
therefore , the total energy that the pulsar has lost over its age , @xmath155 , is @xmath156 .
\label{etot}\ ] ] for the crab pulsar , produced in the supernova of 1054 , a @xmath154 of @xmath157 years is derived from eq .
( [ eqtau0 ] ) , and the total energy is derived as @xmath158ergs with a braking index of @xmath159 from eq .
( [ etot ] ) @xcite . for psr b1800@xmath021 ,
the braking index and the true age are unknown .
the measured braking indices of other pulsars fall in the range of @xmath160 @xcite .
therefore , we estimated the total energies for several braking indices with @xmath161 ergs / s , an age of @xmath162kyears ( the same value as the characteristic age ) , and @xmath163 years ( the same as the crab pulsar ) , which are listed in table [ bindex ] . for the leptonic model , we assumed that the rate of electrons - injection , @xmath164 , proportionally decreases with @xmath151 , that is , @xmath165 , and we considered the cooling effect due to synchrotron and ic energy losses @xcite .
the old electrons that were injected with high - power initial spin - down luminosity have been cooled . on the other hand
the young electrons have less time to be cooled .
figure [ sed_pwnall ] shows the sed with the leptonic model curves with a time - dependent rate of electrons - injection with @xmath163years , a braking index of @xmath166 , an age of @xmath167kyears , @xmath168 , and @xmath1698@xmath131 g : to obtain the model curves , we used the broken energy due to synchrotron cooling and ic cooling , @xmath170 , where @xmath171 is the thomson cross - section and @xmath172 is the total energy density of the photon field . for the pwn of psr b1800@xmath021
, we used the error region for the entire pwn ( inner + outer ) obtained by _
chandra _ @xcite .
the old electrons are responsible for the tev gamma - ray emission , and the young electrons are responsible for the x - ray emission .
this picture may also explain the size difference between the tev gamma - ray and the x - ray emission regions , because the old electrons can extend further than the young ones . to calculate the total energy of electrons , @xmath96 , for the pwn origin scenario
, we substituted @xmath173 for @xmath89 .
then , @xmath96 is estimated to be @xmath174ergs for the above parameters .
an @xmath175 of @xmath176ergs for the same parameters can easily explain the amount of electron energy , and the efficiency of the energy to accelerate electrons is @xmath177% . note that the above parameters are not unique , and other combinations of a @xmath154 and a braking index are possible .
if we adopt @xmath178 years , acceptable model curves can be derived with a braking index of @xmath179 , an age of @xmath167 kyears , @xmath180 , and @xmath181@xmath131 g . for the latter parameters , @xmath175
is estimated to be @xmath182ergs , and @xmath96 is @xmath183ergs .
the efficiency of the energy to accelerate the electrons is @xmath184% . in either case , a quite high value for the total energy , @xmath175 , that the pulsar has lost is required to explain the observed spectra .
we consider the hadronic model given in fig .
[ sed_hadronpwn ] .
table [ summary ] lists the fitting parameters and the total energy of protons @xmath109 at a distance for psr b1800@xmath021 of 3.84 kpc with the assumption that the ism density is @xmath108 @xmath29 .
@xmath109 with @xmath104 of 2.3 is estimated to be @xmath185 kpc)@xmath111 @xmath112ergs . with a @xmath104 of 1.0
, @xmath109 is estimated to be @xmath186 kpc)@xmath111 @xmath112ergs .
the magnetic field is constrained to be @xmath187 mg with the assumption of @xmath188 and the constraint of @xmath189tev ( table [ summary ] ) .
the magnetic field is very high .
@xmath126 is estimated to be more than @xmath190@xmath112 .
if a gmc is associated with hess j1804@xmath0216 , and @xmath114 @xmath29 , the obtained @xmath126 is consistent with that of the average cosmic rays in our galaxy , @xmath135 .
however , the lower limit of the magnetic field is much higher than the typical interstellar magnetic field , and there is no evidence supporting such a high magnetic field in this position . *
suzaku src2 : * for the leptonic model , a similar scenario as that for pwn of psr b1800@xmath021 can be considered for suzaku src2 .
however , because we have no information about the age and the spin - down luminosity of suzaku src2 , we assume the same age and spin - down luminosity as that of psr b1800@xmath021 .
figure [ sed_src2 ] shows leptonic model curves for suzaku src2 with @xmath191 years , a braking index of @xmath192 , an age of @xmath167 kyears , @xmath180 , and b = 8 @xmath131 g .
the total energy of the electrons is calculated to be @xmath193 kpc)@xmath50ergs , and @xmath175 is @xmath194ergs .
this @xmath175 would be proportionally reduced if @xmath195 is lower than that of psr b1800@xmath021 .
for example , @xmath175 is estimated to be @xmath196 ergs with the assumption of @xmath197 ergss@xmath5 .
radio and x - ray observations with higher time resolutions might be able to detect a sub - second pulse period from this source , and further observations in the ir to uv band for a diffuse source in the extended region of hess j1804@xmath0216 could contribute to a tighter solution for the leptonic models .
we consider the hadronic model as given in fig .
[ sed_hadronpwn ] .
the magnetic field is constrained to be @xmath198@xmath131 g with the assumption of @xmath188 and the constraint of @xmath189tev ( table [ summary ] ) .
@xmath126 is estimated to be more than @xmath199@xmath112 .
the lower limit of the magnetic field is much higher than the typical interstellar magnetic field .
lllc vla and other radio & [email protected] & filled circle & ( 1 ) + vla & [email protected] & open circle & ( 2 ) + _ suzaku _ xis & suzaku src1 & closed region ( dashed ) & ( 3 ) + _ suzaku _ xis & suzaku src2 & closed region ( dot - and - short dashed)&(3 ) + _ suzaku _ xis & diffuse ul ( 90% c.l . ) & arrow ( dashed ) & ( 3 ) + _ chandra _ acis & suzaku src2 & closed region ( dot - and - long dashed)&(4 ) + _ chandra _ acis & pwn of psr b1800@xmath021 & closed region ( solid)&(5 ) + _ chandra _ acis & diffuse ul & arrow ( dotted ) & ( 4 ) + _ integral _ ibis & upper limit & arrow ( cross ) & ( 4 ) + _ egret _ & upper limit & arrow ( solid ) & ( 4)(6 ) + _ glast _ & 1 year sensitivity & dash - dot - dotted line & ( 7 ) + 2mass & nir point ul & triangle & ( 8) + h.e.s.s . & hess j1804@xmath0216 & open square & ( 9 ) + cangaroo - iii & hess j1804@xmath0216 & filled square & this work cccc @xmath89 [ tev ] & 3.6 & 2.3 & 1.3 + @xmath96 [ @xmath200 & 330 & 17 & 1.4 + upper limit of b [ @xmath131 g ] & 1 & 8 & 50 [ sum_ic ] ccccc @xmath105 [ tev ] & 30 & 16 & 7.9 & 5.4 + @xmath109[@xmath201 & 9.3 & 3.1 & 1.3 & 0.96 + @xmath109[@xmath201 & 5.9 & 2.0 & 0.84 & 0.61 ccc 3 & @xmath202 & 210 + 2.5 & @xmath203 & 120 + 2.2 & @xmath204 & 72 + 1.8 & @xmath205 & 19 + 1.6 & @xmath206 & 5.2 + 1.4 & @xmath207 & 0.37
cangaroo - iii observed hess j1804@xmath0216 , and detected gamma rays above 600gev at the @xmath1 level during an effective exposure of 76 hr .
the obtained differential flux is consistent with the previous h.e.s.s .
result , and the obtained morphology shows extended emission compared to our point spread function .
we have discussed the radiation mechanism and considered the proposed counterparts to the tev gamma - ray source . for the snr scenario ,
the most plausible counterpart is the snr [email protected] , and both hadronic and leptonic processes can produce the observed tev gamma - ray spectrum .
we obtained upper limits of spectral indices , @xmath208 for leptonic scenario , and @xmath209 for hadronic scenario .
the total energy of the primary electrons , @xmath96 , is estimated to be @xmath210 kpc)@xmath211 ergs with @xmath212 .
the total energy of the protons , @xmath109 , is estimated to be @xmath110 kpc)@xmath111 @xmath112ergs with @xmath213 .
the obtained @xmath109 indicates that a molecular cloud might be associated to satisfy the hadronic scenario for the energetics . extending this to include
an estimation of the emission from secondary electrons will provide a tighter solution for hadronic scenario .
for the pwn scenario , we discussed both leptonic and hadronic processes .
we confirmed that the leptonic model with a time - dependent rate of electron - injection while considering the braking effect for the spin - down luminosity and the cooling effect due to synchrotron and ic energy losses could explain both the high tev gamma - ray flux and the low x - ray flux .
however , a quite high value for the total energy that the pulsar has lost is required for psr b1800@xmath021 in this model , @xmath214ergs . for suzaku src2
, we could obtain an acceptable model curve , and @xmath175 was estimated to be @xmath97ergs with the assumption of @xmath215ergss@xmath5 .
however , we need radio or x - rays observations to determine whether or not there is a pulse period in order to discuss the energetics further .
the model curves that are given in this paper indicate that _ glast _ could determine the spectral index of the accelerated particles , and further observations in the ir to uv band for a diffuse source could give a solution of the counterpart and the radiation mechanism .
the authors would like to thank b. lott for providing us the data of _ glast _ lat sensitivity .
we also thank h. yamaguchi for discussions on snrs .
this work was supported by a grant - in - aid for scientific research by the japan ministry of education , culture , sports , science and technology ( mext ) , the australian research council , and the inter - university research program of the institute for cosmic ray research . the work is also supported by a grant - in - aid for the 21st century center of excellence programs `` center for diversity and universality in physics '' and `` quantum extreme systems and their symmetries '' from mext of japan .
we thank the defense support center woomera and bae systems , and acknowledge all of the developers and collaborators on the galprop project .
y. higashi and t. nakamori were supported by japan society for the promotion of science research fellowships for young scientists .
aharonian , f. a. , et al .
2005a , science , 307 , 1938 aharonian , f. a. , et al .
2005b , a&a , 442 , l25 aharonian , f. a. , et al . 2006a , apj , 636 , 777 aharonian , f. a. , et al .
2006b , a&a , 448 , 43 altenhoff , w. j. , downes , d. , pauls , t. , & schraml , j. 1978 , a&as , 35 , 23 bamba , a. , et al .
2007 , pasj , 59 , s209 blitz , l. , fich , m. , & stark , a. a. 1982 , apjs , 49 , 183 blondin , j. , chevalier , r. , & frierson , d. 2001 , apj , 563 , 806 brisken , w. f. , carrillo - baragan , m. , kurtz , s. , & finley , j. p. 2006 ,
apj , 652 , 554 brogan , c. l. , gelfand , j. d. , gaensler , b. m. , kassim , n. e. , & lazio , t. j. 2006 , apjl , 639 , l25 clifton , t. r. , & lyne , a. g. 1986 , nature , 320 , 43 cordes , j. m. , & lazio , t. j. w. 2002 , astro - ph/0207156 cramphorn , c. k. , & sunyaev , r. a. 2002 , a&a , 389 , 252 cui , w. , & konopelko , a. 2006 , apj , 652 , l109 ( erratum 2007 , apj , 665 , l83 ) enomoto , r. , et al .
2002 , astropart .
phys . , 16 , 235 enomoto , r. , et al . 2006a , apj , 638 , 397 enomoto , r. , et al . 2006b , apj , 652 , 1268 fatuzzo , m. , melia , f. , & crocker , r. m. 2006 , astro - ph/0602330 finley , j. p. , & gelman , h. 1994 , apj , 434 , l25 fisher , r. a. 1936 , annals of eugenics , 7 , 179 frail , d. a. , kassim , n. e. , & weiler , k. w. 1994 , aj , 107 , 1120 funk , s. , hinton , j. a. , phlhofer , g. , aharonian , f. a. , hofmann , w. , reimer , o. , & wagner , s. 2007 , apj , 662 , 517 lat performance 2007 , + http://www-glast.slac.stanford.edu/software/is/glast_lat_performance.htm hamilton , a. j. s. , chevalier , r. a. , & sarazin , c. l. 1983 , apjs , 51 , 115 handa , t. , sofue , y. , nakai , n. , hirabayashi , h. , & inoue , m. 1987 , pasj , 39 , 709 hartman , r. c. , et al . 1999 ,
apjs , 123 , 79 hillas , a. m. 1985 , proc .
19th icrc(la jolla ) , 3 , 445 hinton , j. 2007 , arxiv.0712.3352 kabuki , s. , et al .
2003 , nucl.inst.meth .
, a500 , 318 kabuki , s. , et al .
2007 , apj , 668 , 968 kargaltsev , o. , pavlov , g. g. , & garmire , g. p. 2007a
, apj , 660 , 1413 kargaltsev , o. , pavlov , g. g. , & garmire , g. p. 2007b
, apj , 670 , 643 kassim , n. e. , & weiler , k. w. 1990a , nature , 343 , 146 kassim , n. e. , & weiler , k. w. 1990b , apj,360,184 kawachi , a. , et al .
2001 , astropart .
, 14 , 261 kubo , h. , et al .
2001 , proc .
27th icrc(hamburg ) , 2900 landi , r. , et al .
2006 , apj , 651 , 190 livingstone , m. a. , kaspi , v. m. , gotthelf , e. v. , & kuiper , l. 2006 , apj , 647 , 1286 mazzali , p. a. , et al .
2006 , apj , 645 , 1323 nishijima , k. , et al .
2005 , proc .
29th icrc(pune ) , 5 , 327 odegard , n. 1986 , aj , 92 , 1372 ojeda - may , p. , kurts , s. e. , rodrguez , l. f. , arthur , s. j. , & velzquez , p. f. 2002 , rev.mexicana.astrofis . ,
38 , 111 pacini , f. , & salvati , m. 1973 , apj , 186 , 249 pfrommer , c. , enlin , t. a. , 2004 , a&a , 413 , 17 porter , t. a. , & strong , a. w. 2005 , proc .
29th icrc(pune ) , 4 , 77 reich , w. , frst , e. , steffen , p. , reif , k. , & haslam , c. g. t. 1984 , a&as , 58 , 197 sedov , l. i. 1959 , similarity and dimensional methods in mechanics , new york ; academic press skrutskie et al .
2006 , aj , 131 , 1163 strong , w. s. , & moskalenko , i. v. 2006 , galprop project , http://galprop.stanford.edu/web_galprop/galprop_home.html taylor , j. h. , & cordes , j. m. 1993 , apj , 411 , 674 white , r. l. , becker , r. h. , & helfand , d. j. 2005 , aj , 130 , 586 yamazaki , r. , kohri , k. , bamba , a. , yoshida , t. , tsuribe , t. , & takahara , f. 2006 , mnras , 371 , 1975 | we observed the unidentified tev gamma - ray source hess j1804@xmath0216 with the cangaroo - iii atmospheric cerenkov telescopes from may to july in 2006 .
we detected very high energy gamma rays above 600 gev at the @xmath1 level in an effective exposure of 76 hr .
we obtained a differential flux of @xmath2(e/1 tev)@xmath3 @xmath4s@xmath5tev@xmath5 with a photon index @xmath6 of @xmath7 , which is consistent with that of the h.e.s.s .
observation in 2004 .
we also confirm the extended morphology of the source . by combining our result with multi - wavelength observations ,
we discuss the possible counterparts of hess j1804@xmath0216 and the radiation mechanism based on leptonic and hadronic processes for a supernova remnant and a pulsar wind nebula . |
our first goal is to build a lattice which is integrable by construction . for the sake of simplicity ,
we consider first a 2-dimensional lattice @xmath8 . given the three arbitrary points @xmath9 , the forth point @xmath10 is fixed prescribing a rule ( the one defining the lattice ) ; since we want an _ integrable lattice _ , this _ rule must be linear_. we shall impose the simplest linear rule : _ planarity _ ; i.e. , the forth point @xmath10 will belong to the plane generated by the first three points @xmath11 and @xmath12 or , in algebraic terms , @xmath13 where @xmath14 is the translation operator in the @xmath15 direction and @xmath16 is the corresponding difference operator .
given two discrete initial curves ( two sequences of points ) @xmath17 , @xmath18 and two arbitrary sets of functions @xmath19 , the planarity constraint allows one to construct uniquely a two - dimensional lattice , which we have called _ quadrilateral _ ( or _ planar _ ) .
the quadrilateral surface was first proposed in @xcite as the proper discrete analogue of a conjugate net on a surface , without any connection with integrability . indeed , at the moment
, the above construction does not seem to have anything to do with integrability ; to show its profound connection with the familiar integrability schemes it is necessary to generalize the picture to higher dimensions , imposing the planarity constraint on each two - dimensional @xmath20 surface of the lattice @xcite .
an @xmath21 dimensional lattice @xmath22 is a quadrilateral lattice ( ql ) iff its elementary quadrilaterals @xmath23 are planar ; i.e. , iff @xmath24 the planarity constraints ( [ eq : laplace ] ) are compatible only for the special class of data @xmath25 satisfying the nonlinear system @xmath26 these equations characterize the quadrilateral lattice and we refer to them as the ql equations .
they were first derived in @xcite as integrable discrete analogues of the darboux equations for conjugate nets , but without any geometric characterization .
we remark that , if a quadrilateral lattice exists , it is integrable by construction , since it is built out of the linear constraints ( [ eq : laplace ] ) .
indeed , in the language of the theory of integrable systems , the planarity constraints correspond to the set of linear spectral problems ( [ eq : laplace ] ) and the resulting ql and its characterizing equations ( [ eq : mql - a ] ) correspond to the integrable nonlinear equations arising as the compatibility condition for such spectral problems .
actually , in connection with integrability , certain constructions based on the quadrilateral surfaces were already known : the integrable discrete analogue of isothermic surfaces @xcite and the laplace sequence of quadrilateral surfaces @xcite , which provides the geometric meaning for the hirota equation ( see also section [ sec : hir ] and @xcite for more details ) .
it is often convenient to reformulate equations ( [ eq : laplace ] ) as a first order system @xcite .
to do so we introduce the suitably scaled tangent vectors @xmath27 , @xmath28 : @xmath29 in such a way that the @xmath30-th variation of @xmath27 is proportional to @xmath31 only : @xmath32 figure 1 .
the compatibility condition for the system ( [ eq : lin - x ] ) gives the following new form of the ql equations @xmath33 and the scaling factors @xmath34 , called the lam coefficients , solve the linear equations @xmath35 whose compatibility gives equations ( [ eq : mql - q ] ) again ; moreover @xmath36 . the solution of the initial value problem for the ql is contained in the following result @xcite .
given @xmath21 initial discrete curves @xmath37 , @xmath38 , assigning the initial data @xmath39 , @xmath40 ( or , equivalently , the initial data @xmath41 ) , one constructs the initial @xmath42 quadrilateral surfaces of the lattice applying equations ( [ eq : laplace ] ) ( or , equivalently , equations ( [ eq : lin - x ] ) ) .
then the quadrilateral lattice @xmath43 follows uniquely from the planarity constraint .
figure 2 . in the continuous limit
: @xmath44 the ql reduces to an @xmath21 dimensional conjugate net in @xmath45 , characterized by the famous darboux equations @xcite @xmath46 in this section we define the backward data @xmath47 , @xmath48 , @xmath49 of the quadrilateral lattice .
it turns out that the relation between the forward data @xmath27 , @xmath34 , @xmath50 and the backward data is given in terms of the @xmath0function @xcite , which is one of central objects of the theory of integrable systems . in the previous section
the quadrilateral lattice was built through a _
forward construction _
; it is of course possible to build the lattice also through a _
backward construction_. the backward tangent vectors @xmath47 and the backward lam coefficients @xmath48 , @xmath51 are defined by the equations @xmath52 in terms of the backward difference operator @xmath53 .
the backward lam coefficients are again chosen in such a way that the @xmath54 variation of @xmath55 is proportional to @xmath47 only .
we define the backward rotation coefficients @xmath49 as the corresponding proportionality factors @xmath56 figure 3 . comparing equations ( [ eq : lin - h ] ) and ( [ eq : lin - bx ] ) we see immediately that the new functions @xmath49 satisfy the mql equations ( [ eq : mql - q ] ) as well .
moreover the new scaling factors @xmath48 satisfy the following system of linear equations @xmath57 whose compatibility condition gives again the ql equations ( [ eq : mql - q ] ) . an easy consequence of equations ( [ eq : b - h - x ] ) , ( [ eq : lin - bx ] ) and ( [ eq : lin - bh ] ) is the following the vector function @xmath22 representing a quadrilateral lattice satisfies the backward laplace equation @xmath58 where , in the notation of this section , @xmath59 .
the forward and backward rotation coefficients @xmath50 and @xmath49 describe the same lattice @xmath43 from different points of view , therefore they are not independent . indeed , defining the functions @xmath60 as the proportionality factors between @xmath27 and @xmath61 ( both vectors are proportional to @xmath62 ) : @xmath63 we have the following the forward and backward data of the lattice @xmath43 are related through the following formulas @xmath64 and the factors @xmath65 are first potentials satisfying equations @xmath66 since @xmath50 and @xmath49 are both solutions of the ql equations ( [ eq : mql - q ] ) , then equations ( [ eq : q - qt ] ) , ( [ eq : rho - constr ] ) describe a special symmetry transformation of equations ( [ eq : q - qt ] ) .
the rhs of equation ( [ eq : rho - constr ] ) is symmetric with respect to the interchange of @xmath15 and @xmath30 ; this implies the existence of a potential @xmath67 , such that @xmath68 therefore equation ( [ eq : rho - constr ] ) defines the second potential @xmath0 : @xmath69 the potential @xmath0 connecting the forward and backward data : @xmath70 is the famous @xmath0-_function _ of the quadrilateral lattice . in terms of the @xmath0 function ,
the ql equations read as follows @xcite : @xmath71 @xmath72 where @xmath73
we present the basic ideas and results of the theory of the darboux type transformations of the multidimensional quadrilateral lattice .
our approach follows the spirit of the book @xcite , which summarizes the theory of transformations of ( two dimensional ) conjugate nets .
we also name the transformations of the quadrilateral lattice according to the classical geometric terminology of the transformations of conjugate nets .
all the results of this section can be found in @xcite , to whom we also refer for proofs and more detailed explanation . to define the transformations we need another geometric object
the _ congruence _ which serves as the link between two quadrilateral lattices .
[ def : int - congr ] an @xmath21-dimensional _ rectilinear congruence _
( or , simply , congruence ) is a mapping @xmath74 from the integer lattice to the space of lines in @xmath45 such that every two neighboring lines @xmath75 and @xmath76 , @xmath28 , are coplanar .
we remark that with any @xmath21 dimensional quadrilateral lattices we may naturally associate @xmath21 congruences as follows .
[ def : tan - congruence ] given an @xmath21-dimensional quadrilateral lattice @xmath43 , its @xmath15-th _ tangent congruence _
@xmath77 consists of the lines passing through the points @xmath43 of the lattice and directed along the tangent vectors @xmath62 . reversing that process , one can associate with any @xmath21-dimensional congruence @xmath21 lattices which are , in general , quadrilateral lattices .
the @xmath15-th _ focal lattice _ @xmath78 of a congruence @xmath75 is the lattice constructed out of the intersection points of the lines @xmath79 with @xmath80 .
there exists a simple ( but basic for the theory of transformations ) relation between quadrilateral lattices and congruences .
[ def : conj - congr - latt ] a quadrilateral lattice and a congruence are called _ conjugate _ if there exists a one - to - one correspondence between points of the lattice and lines of the congruence such that lines pass through the corresponding points .
focal lattices of congruences conjugate to quadrilateral lattices are quadrilateral lattices .
the above introduced notions allow to define all the transformations of quadrilateral lattices .
[ def : fundamental ] two quadrilateral lattices are related by the
_ fundamental transformation @xmath81 of jonas _ when they are conjugate to the same congruence , which is called the congruence of the transformation .
usually one assumes that , in the fundamental transfomation , the congruence is _ transversal _ to the lattice , i.e. , non - tangent to the lattice in corresponding points .
otherwise we get the following reductions of the fundamental transformation * the congruence of the transformation is the @xmath15-th tangent congruence of the original lattice the _ lvy transformation _ @xmath82 * the congruence of the transformation is the @xmath15-th tangent congruence of the new lattice the _ adjoint lvy transformation _
@xmath83 * both lattices are focal lattices of the congruence of the transformation
the _ laplace transformation _
figure 4 .
in addition , there are two useful special transformations * the _ projective _ or _ radial transformation _ , when all the lines of the congruence meet in a single point * the _ combescure transformation _
, when the transformed lattice is parallel to the original one let us present some formulas for the transformations , the details can be found in @xcite .
[ th : fund ] the fundamental transformation @xmath84 of the quadrilateral lattice @xmath43 is given by @xmath85 where + i ) @xmath86 is a new solution of the laplace equation ( [ eq : laplace ] ) of the lattice @xmath43 + ii ) @xmath87 is the combescure transformation vector , which is a solution of the equations @xmath88 where , due to the compatibility of the system ( [ eq : dixc ] ) , the functions @xmath89 satisfy @xmath90 moreover + iii ) @xmath91 is a solution , corresponding to @xmath92 , of the laplace equation of the lattice @xmath87 , i.e. @xmath93 the combescure transformation vector @xmath87 is used to construct the congruence of the transformation , and the function @xmath92 serves to place points of the new lattice on the lines of the congruence . the definitions ( [ eq : fund ] ) and ( [ eq : dixc])-([eq : djsi ] ) of the fundamental and combescure transformations were independently presented in @xcite . notice that , given @xmath87 and @xmath92 , then equation ( [ eq : diphc ] ) determines @xmath91 uniquely , up to a constant of integration .
one can stop the transformation of the lattice @xmath43 at the intermediate level of construction of the vector @xmath87 to obtain the combescure transformation @xmath94 characterized by the property that the tangent lines to both lattices in corresponding points are parallel .
[ def : levy ] the @xmath15-th _ lvy transform _
@xmath95 of the quadrilateral lattice @xmath43 is a quadrilateral lattice conjugate to the @xmath15-th tangent congruence of @xmath43 .
since the congruence of the transformation is known , to construct the transformation we need only the solution @xmath92 of the laplace equation of the lattice @xmath43 .
[ prop : levy ] the lvy transform @xmath95 of the quadrilateral lattice @xmath43 is given by @xmath96 another way to reduce the fundamental transformation is to fix the way we place the new lattice on the congruence .
[ def : adj - levy ] the @xmath15-th _ adjoint lvy transform _ @xmath97 of the quadrilateral lattice @xmath43 is the @xmath15-th focal lattice of a congruence conjugate to @xmath43 . to find the adjoint lvy transformation we need only the congruence of the transformation , which can be find via the combescure transformation vector @xmath87
the adjoint lvy transform of the lattice @xmath43 is given by @xmath98 where the functions @xmath89 are defined in theorem [ th : fund ] .
the deepest reduction of the fundamental transformation is the laplace transformation where both the congruence of the transformation and the way to place the new lattice on the congruence are fixed .
the _ laplace transform _
@xmath99 of the quadrilateral lattice @xmath43 is the @xmath30-th focal lattice of its @xmath15-th tangent congruence .
figure 5 .
[ prop : laplace ] the laplace transformation of the quadrilateral lattice @xmath43 is given by @xmath100 the superpositions of laplace transformations satisfy the following identities @xmath101 finally , the notion of radial ( or projective ) transformation is related to the so called radial congruence , whose focal lattices are degenerated to a single point , which can be taken as the origin . the radial ( or projective ) transform @xmath102 of the quadrilateral lattice @xmath43 is a quadrilateral lattice conjugate to the radial congruence of @xmath43 .
[ prop : radial ] the radial transform @xmath102 is given by @xmath103 where @xmath86 is a solution of the laplace equation ( [ eq : laplace ] ) of the lattice @xmath43 .
the degeneration of the fundamental transformation to the lvy , adjoint lvy , the combescure , radial , and the laplace transformations can be shown also on the analytic level . from another point of view , the fundamental transformation can be decomposed into the superposition of the lvy and adjoint lvy transformations , or into the superposition of two combescure and one radial transformation .
moreover the fundamental transformation can be put into the scheme of the laplace transformation ( preceeded by building a new level on the original lattice ) .
the laplace transformations of two dimensional conjugate nets generates the two dimensional toda system @xcite .
analogously , one can associate @xcite ( see also @xcite for more datails and new interpretations ) the integrable discrete version of the toda system ( the hirota equation @xcite ) with the laplace transformations of a two dimensional quadrilateral lattice . given a quadrilateral surface @xmath104 , consider the sequence @xmath105 of quadrilateral surfaces obtained from @xmath43 by resursive application of the laplace transformation @xmath106 : @xmath107 the above sequence , which is the proper analogue of the laplace sequence of conjugate nets , is well defined due to formula ( [ eq : lid - ij ] ) .
the coefficients @xmath108 , @xmath109 of the laplace equations of the sequence of lattices satisfy the following system @xmath110 notice that one can consider the parameter @xmath111 along the laplace sequence as the third discrete variable .
actually , the laplace sequence of quadrilateral surfaces may be viewed as a deep reduction of the three dimensional quadrilateral lattice in the same sense like the laplace transformations are reductions of the fundamental transformation .
an elementary quadrilateral of such reduced ( or degenerated ) lattice with @xmath112 is shown below .
figure 6 .
define the function @xmath113 as the cross - ratio of four collinear points : @xmath114 , @xmath115 , @xmath116 and @xmath117 .
simple derivation shows that @xmath118 and equations ( [ eq : todaa1])-([eq : todaa2 ] ) imply that @xmath119 satisfies the gauge invariant form of the hirota equation @xmath120 consider @xmath121 fundamental transformations @xmath122 , @xmath123 , of the quadrilateral lattice @xmath43 , which are built from + i ) @xmath119 solutions @xmath124 , @xmath123 of the laplace equation of the lattice @xmath43 ; + ii ) @xmath119 combescure transformation vectors @xmath125 , where @xmath126 and @xmath127 satisfy equations @xmath128 iii ) @xmath119 functions @xmath129 , which satisfy @xmath130 we arrange functions @xmath124 in the @xmath119 component vector @xmath131 ; similarily , we arrange the combescure transformation vectors @xmath125 into the @xmath132 matrix @xmath133 ; moreover we introduce + iv ) the @xmath134 matrix @xmath135 , whose columns are the @xmath119 component vectors @xmath136 , which are the combescure transforms of @xmath137 @xmath138 the diagonal part of @xmath139 is fixed by the initial fundamental transformations . to find the off - diagonal part of @xmath139 we integrate equations ( [ eq : int - phic ] ) introducing @xmath140 arbitrary constants .
one can show that the vectorial fundamental transformation @xmath141 of the quadrilateral lattice @xmath43 , which is defined as @xmath142 is again a quadrilateral lattice .
moreover , the vectorial transformation is the superposition of the fundamental transformations @xmath143 and does not depend on the order in which the transformations are taken . in applying the fundamental transformations at the intermediate stages ,
the transformation data should be suitably transformed as well . to prove the superposition and permutability statements , it is important to notice the following basic fact .
[ lem : sup - vect - fund ] assume the following splitting of the data of the vectorial fundamental transformation @xmath144 associated with the partition @xmath145 .
then the vectorial fundamental transformation @xmath141 is equivalent to the following superposition of vectorial fundamental transformations : + 1 .
transformation @xmath146 with the data @xmath147 , @xmath148 , @xmath149 : @xmath150 2 .
application on the result obtained in point 1 .
, transformation @xmath151 with the data transformed by the transformation @xmath152 as well @xmath153 where @xmath154 another important feature of the fundamental transformation is that its recursive application generates new dimensions of the quadrilateral lattice . in particular , when applied to the continuous limit of the quadrilateral lattice
the conjugate net this procedure explains , on the geometric level , the power of using the darboux type transformations in discretizing the integrable equations .
figure 7 .
the continuous limit of the above theory reduces to the classical theory of finite transformations of @xmath21 dimensional conjugate nets .
it was recently shown in @xcite that these transformations of conjugate nets are strictly related to the quantum field theoretical formulation of the multicomponent kadomtsev petviashvilii hierarchy .
the @xmath155-@xmath156 system of quantum fields , which appears as the system of ghost fields in string theory , is constructed in terms of the anticommutation relations @xmath157 where @xmath158 and @xmath159 , @xmath160 , are free charged fermion fields defined on the unit circle @xmath161 , and @xmath162 is the dirac distribution on @xmath161 . the clifford algebra generated by the @xmath155-@xmath156 system admits a representation in terms of bosonic variables . in this representation
the fields act on the fock space @xmath163 of complex - valued functions @xmath164 with @xmath165 the representation of the @xmath155-@xmath156 generators takes the form @xcite : @xmath166{{\bf e}}_i),\\ c_i(z)\tau({{\bf \ell}},{{\bf t}})&=&(-1)^{\sum_{j > i}\ell_j }
z^{-\ell_i}\exp(-\xi(z,{{\bf t}}_i ) ) \tau({{\bf \ell}}+{{\bf e}}_i,{{\bf t}}+\left [ 1/z \right]{{\bf e}}_i),\end{aligned}\ ] ] where @xmath167 : = \left ( \frac{1}{z } , \frac{1}{2z^2 } , \frac{1}{3z^3 } , \ldots \right),\ ] ] and @xmath168 are the canonical generators of @xmath169 .
the fock space decomposes into a direct sum of charge sectors @xmath170 where the total charge operator @xmath171 is the sum of @xmath21 commuting partial charges @xmath172 , @xmath173 ; they correspond to the @xmath21 different flavours of fermions of the model .
the @xmath21-component kp hierarchy can be formulated as the following bilinear identity @xmath174 where @xmath175 in terms of the components @xmath176 we have @xmath177{{\bf e}}_m ) \tau({{\bf \ell}}^\prime+{{\bf e}}_m,{{\bf t}}^\prime+\left [ 1/z \right]{{\bf e}}_m)=0 & & \nonumber,\end{aligned}\ ] ] for any @xmath178 , @xmath179 and @xmath180 , @xmath181 such that @xmath182 .
denote by @xmath183 the shift operators @xmath184 correspond to the so called schlesinger transformations in monodromy theory @xcite and do not alter the neutral character of the assembly of fermions .
moreover , @xmath184 are obvious symmetries of ( [ bilineal ] ) . the @xmath185 matrix baker function @xmath186 and its adjoint @xmath187 can be defined in terms of the @xmath0 function as @xmath188{{\bf e}}_j)}{\tau({{\bf t}})}\exp(\xi(z,{{\bf t}}_j ) ) , \\ \psi_{ij}^*(z,{{\bf t}})&=&\varepsilon_{ji}z^{\delta_{ij}-1}\exp(-\xi(z,{{\bf t}}_i ) ) \frac{\tau_{{ij}}({{\bf t}}+\left [ 1/z \right]{{\bf e}}_i)}{\tau({{\bf t}})},\end{aligned}\ ] ] where @xmath189 , @xmath190 ( @xmath191 ) .
the connection of the @xmath21 component hierarchy with the conjugate nets is given in the following result @xcite [ cor : mkp - nets ] the solutions of the n - component kp hierarchy describe @xmath21 dimensional conjugate nets with coordinates @xmath192 , @xmath51 , while the remaning times @xmath193 , for @xmath194 , describe integrable iso - conjugate deformations of the nets .
in particular , the rotation coefficients are given by @xmath195 the tangent vectors @xmath27 are constructed from the rows of the matrix @xmath196 for some distribution matrix @xmath197 ; the lam coefficients are given by the entries of the row matrix @xmath198 for some distribution row matrix @xmath199 .
as one can expect , various elements of the kp theory have their conterparts on the level of the conjugate nets .
\i ) the schlesinger transformation @xmath184 gives rise to the laplace transformation @xmath200 of the conjugate nets .
+ ii ) the action of the ( vertex ) operators @xmath201 and @xmath202 gives rise to the lvy and the adjoint lvy transformations @xmath82 and @xmath83 .
correspondingly + iii ) the soliton vertex operator generated by the superposition @xmath203 gives rise to the fundamental transformation .
in this section we introduce and study the properties of quadrilateral lattices in the dual space ( hyperplane lattices ) @xcite .
these considerations turn out to be relevant in the reduction theory of the quadrilateral lattices , when the introduction of the additional geometric structure will allow to establish a direct connection between point lattices and hyperplane lattices .
generic hyperplanes can be represented by the co - vectors @xmath204 and consist of points @xmath205 satisfying the equation @xmath206 the equation of the hyperplane passing through @xmath207 and parallel to that represented by @xmath208 can be written as @xmath209 ( the co - vector @xmath210 represents the hyperplane at infinity ) . the linear mappings between points and hyperplanes are called _
correlations_. given the sphere of radius @xmath211 centered at the orgin , it defines the corresponding corelation @xmath212 ( called _ polarity _ with respect to that sphere ) .
the image @xmath213 of a point @xmath214 is called the _ polar hyperplane _ of @xmath215 , and consists of points @xmath43 satisfying equation @xmath216 , i.e , @xmath217 . the notions dual to the parallelism of two lines and to the planarity of four points are contained in the following [ def : co - parallelity ] i ) two subspaces of co - dimension @xmath218 are called `` co - parallel '' if there exists a hyperplane passing through the origin and containing them .
\ii ) four hyperplanes are `` co - planar '' if the fourth hyperplane contains the intersection of the first three . a _ hyperplane lattice _
is simply a lattice @xmath219 , @xmath220 , in the space of hyperplanes . from definition [ def : co - parallelity ]
we obtain the following geometric object dual to the quadrilateral lattice . the hyperplane lattice @xmath221 is `` quadrilateral '' ( `` co - planar '' ) if , for any @xmath222 , @xmath223 , the hyperplane @xmath224 contains the subspace @xmath225 .
this property reduces again to the laplace equation : @xmath226 a convenient way to construct quadrilateral hyperplane lattices makes use of systems of parallel quadrilateral point lattices .
we recall that two quadrilateral lattices @xmath43 , @xmath227 are parallel ( comberscure related ) iff @xmath228 ; in this case the scaled tangent vectors of the two lattices can be chosen to be equal , then the rotation coefficients @xmath229 of both lattices coincide and the lam coefficients @xmath34 and @xmath230 are solutions of the same equation .
consider a system of @xmath231 parallel and linearly independent point lattices @xmath232 , @xmath233 .
denote by @xmath234 the hyperplane lattice made out of the hyperplanes passing through the ( corresponding ) points of @xmath232 and spanned by the vectors @xmath235 , @xmath236 , i.e. , @xmath237 we call the system @xmath238 of hyperplane lattices the `` dual system '' to the system of parallel point lattices @xmath239 .
if we arrange the parallel system @xmath232 , @xmath233 of point lattices as columns of the _ matrix @xmath240 of the system _ :
@xmath241 then we have the following results . 1 .
the dual system @xmath238 consists of co - parallel quadrilateral hyperplane lattices and is given by the rows of the matrix @xmath242 .
2 . the rows of the matrix @xmath240 define a system @xmath243 of co - parallel quadrilateral hyperplane lattices , which we call the `` adjoint system '' to @xmath232 , characterized by the property that the forward rotation coefficients of the system @xmath239 are the backward rotation coefficients of the system @xmath243 : @xmath244 .
we now study some geometrically distinguished reductions of the ql which possess additional geometric properties that , once imposed on the initial surfaces , `` propagate '' everywhere through the construction of the lattice .
since the quadrilateral lattice is integrable , these reductions will inherit its integrability properties . 1 .
a quadrilateral lattice is symmetric iff , for a given set of rotation coefficients @xmath50 , there exists a @xmath0function of the lattice such that @xmath246 or equivalently , in terms of the corresponding potentials @xmath65 , @xmath247 2 . a quadrilateral lattice is symmetric iff , for a given set of forward tangent vectors @xmath27 of the lattice , there exists a complementary set of the backward tangent vectors @xmath47 such that the parallelograms @xmath248 and @xmath249 are similar .
3 . a quadrilateral lattice is symmetric iff , for different indices @xmath250 , its rotation coefficients satisfy the following constraint @xmath251 as we have anticipated , the constraints discussed in this paper allow one to establish a connection between quadrilateral point lattices and their duals , the quadrilateral hyperplane lattices .
the following proposition describes this connection in the case of the symmetry constraint .
[ prop : om - symm ] given a system of parallel quadrilateral lattices @xmath252 and the associated matrix @xmath240 , then the following properties are equivalent . + i ) the matrix @xmath240 of the system is symmetric : @xmath253 ii ) the polar hyperplane @xmath254 of the point lattice @xmath232 coincides with the hyperplane lattice @xmath255 : @xmath256
iii ) the lattices @xmath232 , @xmath233 are symmetric .
furthermore the associated tangent vectors @xmath27 and @xmath257 are related in the following way @xmath258 in the continuous limit ( [ eq : limit ] ) the symmetric quadrilateral lattice reduces to a _ symmetric conjugate net _ , for which the rotation coefficients @xmath259
satisfying the darboux equations ( [ eq : darboux ] ) are symmetric @xmath260 this notion was first proposed in @xcite for @xmath261 , @xmath262 , as a discrete analogue of surfaces parametrized by curvature lines ( see also @xcite ) ; later in @xcite for @xmath263 and , finally , for arbitrary @xmath220 in @xcite ; subsequently a convenient set of equations characterizing the circular lattices in @xmath264 was found in @xcite . a geometric proof of the integrability of the circular lattice was first given in @xcite while the analytic proof of its integrability was given in @xcite through the @xmath265 method .
an elementary characterization of circular quadrilaterals states that , if a circular quadrilateral is convex , then the sum of its opposite angles is @xmath266 ; when the quadrilateral is skew , then its opposite angles are equal .
other convenient characterizations are the following . 1 .
a quadrilateral lattice is circular if and only if : @xmath267 2 .
a quadrilateral lattice is circular if and only if @xcite : @xmath268 3 .
a quadrilateral lattice @xmath43 is circular iff the scalars @xmath269 solve the linear system ( [ eq : lin - x ] ) or , equivalently , iff the function @xmath270 ( the square of the norm of @xmath43 ) satisfies the laplace equation ( [ eq : laplace ] ) of @xmath43 .
this characterization was found in @xcite and explained geometrically in @xcite .
in addition , the circularity constraint ( [ eq : circularity1 ] ) implies the following formula @xcite : @xmath271 which , compared with equations ( [ eq : rho - constr])-([eq : tau ] ) , allows to fix , without loss of generality , the backward formulation of the circular lattice in the following way : latexmath:[\[\label{eq : circularity3 } a distinguished sub - class of circular lattices corresponds to the particular case in which the lattice points @xmath43 belong to the sphere of radius @xmath211 : @xmath273 . in this case there
exists , like for the symmetric reduction , an elegant relation between point lattices and hyperplane lattices @xcite .
[ prop : om - circ ] given a system of parallel quadrilateral lattices @xmath252 and the associated matrix @xmath240 of the system , then the following properties are equivalent . + i ) the matrix @xmath274 is orthogonal : @xmath275 ii ) the polar hyperplane @xmath276 coincides with the dual hyperplane @xmath277 : @xmath278 iii ) the quadrilateral lattices @xmath279 form an orthonormal basis : @xmath280 in addition , the associated tangent vectors @xmath281 , @xmath28 , are related by the following formulas @xmath282 with @xmath283 ) and its adjoint @xmath284 in the previous sections we introduced two integrable reductions of the quadrilateral lattice : the symmetric and the circular lattice . imposing both properties on the initial surfaces of the lattice , they propagate through the lattice in the unique construction determined by the planarity condition and the corresponding lattice will be an integrable deeper reduction of the quadrilateral lattice .
we call this lattice a _ egorov lattice _ since , in the continuous limit , it reduces to a egorov system of coordinates on submanifolds @xcite . if @xmath289 , then it turns out that the egorov lattice is a circular lattice invariant under translations along the main diagonal @xcite , i.e. , such that : @xmath290 where @xmath291 is the total shift : @xmath292 . the continuous limits of the above equations : @xmath296 characterize submanifolds parametrized by egorov systems of conjugate coordinates ( egorov nets ) .
at last , the invariance property ( [ eq : d - inv ] ) reduces to @xmath297 implying that @xmath298 . for @xmath299 ,
we recover a classical characterization of the egorov net @xcite .
quadrilateral lattices in quadrics generalize ( in a sense which will be explained in proposition [ prop : st - clat ] ) the circular lattices .
their continuous analogs the so called _ quadratic conjugate nets _
( see @xcite and references therein ) were the subject of intensive study by ribaucour , darboux , bianchi and eisenhart .
the particular type of fundamental transformation of conjugate nets which preserves the quadratic constraint is called the ribaucour transformation . in this section
we present the basic geometric results concerning quadrilateral lattices in quadrics ( see @xcite for details ) ; in particular , we give the analogue of the ribaucour transformation .
as it was shown in @xcite , such a constraint is compatible with the geometric integrability scheme of the quadrilateral lattice .
this integrability statement can be generalized to quadrilateral lattices in spaces obtained by intersection of many quadric hypersurfaces like , for example , lattices in the grassmann manifolds and in segr or veronese varieties . to show a simple example ,
consider a quadrilateral lattice in the standard @xmath231 dimensional sphere @xmath304 .
the intersection of the plane of any elementary quadrilateral of the lattice @xmath305 with the sphere @xmath304 is a circle . in the stereographic projection
@xmath306 circles of the sphere @xmath304 are mapped into circles ( or straight lines , i.e. , circles passing through the infinity point ) of @xmath307 , therefore we have : [ prop : st - clat ] quadrilateral lattices in the sphere @xmath304 are mapped during the stereographic projection into multidimensional circular lattices in @xmath307 ; conversely , any circular lattice in @xmath307 can be obtained in this way .
theorem [ th : fund ] states that , in order to construct the fundamental transformation of the lattice @xmath43 , we need three new ingredients : @xmath92 , @xmath87 and @xmath91 . in looking for the ribaucour reduction of the fundamental transformation
, we can use the following additional information : + i ) the initial lattice @xmath43 satisfies the quadratic constraint ( [ eq : quadric ] ) , + ii ) the final lattice @xmath308 should satisfy the same constraint as well .
+ this should allow to reduce the number of the necessary data and , indeed , to find the ribaucour transformation , it is enough to know the combescure transformation vector @xmath87 only .
the ribaucour reduction @xmath308 of the fundamental transformation of the quadrilateral lattice @xmath43 subjected to quadratic constraint ( [ eq : quadric ] ) is determined by the combescure transformation vector @xmath87 , provided that @xmath87 is not annihilated by the bilinear form @xmath302 of the constraint @xmath309 the functions @xmath92 and @xmath91 entering in formula ( [ eq : fund ] ) are then given by @xmath310 as it was shown in @xcite , the ribaucour transformation for circular lattices in @xmath307 , defined algebraically in @xcite , can be obtained combining the stereographic projection with the ribaucour transformation of the quadrilateral lattices in @xmath304 . to present the ribaucour reduction of the vectorial fundamental transformation we use results and notation of section 2 , together with equations ( [ def : phi - q ] ) and ( [ def : phic - q ] ) , to obtain @xmath311 equations ( [ eq : int - phic ] ) and ( [ def : phi - qv ] ) lead to @xmath312 which implies that @xmath313 the constant of integration was found from the condition @xmath314 .
[ prop : vect - rib ] the vectorial ribaucour transformation @xmath315 , i.e. , the reduction of the vectorial fundamental transformation ( [ eq : vect - fund ] ) compatible with the quadratic constraint ( [ eq : quadric ] ) , is given by the following constraints @xmath316 [ prop : sup - vect - rib ] assume the following splitting of the data of the vectorial ribaucour transformation @xmath317 associated with the partition @xmath145 .
then the vectorial ribaucour transformation @xmath318 is equivalent to the following superposition of vectorial ribaucour transformations : + 1 .
transformation @xmath319 with the data @xmath147 , @xmath148 , @xmath149 .
application on the result obtained in point 1 .
the transformation @xmath320 with the data @xmath321 , @xmath322 , @xmath323 given by @xmath324-analogs of formulas ( [ eq : f1x2])([eq : f1p22 ] ) .
combining the stereographic projection with the vectorial ribaucour transformation of the quadrilateral lattices in @xmath304 , one can obtain the vectorial ribaucour transformation for circular lattices in @xmath307 and prove its permutability property .
this result was also found independently in @xcite , directly on the level of circular lattices in @xmath307 . e. date and m. kashiwara and m. jimbo and t. miwa , _ transformation groups for soliton equations _ ,
[ in : ] _ nonlinear integrable systems classical theory and quantum theory , proc . of rims
, m. jimbo and t. miwa ( eds . ) , world scientific , 1983 , singapore , pp . 39 - 119 . | we review recent results on integrable discrete geometry .
it turns out that most of the known ( continuous and/or discrete ) integrable systems are particular symmetries of the quadrilateral lattice , a multidimensional lattice characterized by the planarity of its elementary quadrilaterals .
therefore the linear property of planarity seems to be a basic geometric property underlying integrability .
we solve the initial value problem for the quadrilateral lattice and we present the geometric meaning of its @xmath0function , as the potential connecting its forward and backward data .
we present the theory of transformations of the quadrilateral lattice , which is based on the discrete analogue of the theory of rectilinear congruences .
in particular , we discuss the discrete analogues of the laplace , combescure , lvy , radial and fundamental transformations and their interrelations .
we also show how the sequence of laplace transformations of a quadrilateral surface is described by the discrete toda system in three dimensions .
we finally show that these classical transformations are strictly related to the basic operators associated with the quantum field theoretical formulation of the multicomponent kadomtsev petviashvilii hierarchy .
we review the properties of quadrilateral hyperplane lattices , which play an interesting role in the reduction theory , when the introduction of additional geometric structures allows to establish a connection between point and hyperplane lattices .
we present and fully characterize some geometrically distinguished reductions of the quadrilateral lattice , like the symmetric , circular and egorov lattices ; we review also basic geometric results of the theory of quadrilateral lattices in quadrics , and the corresponding analogue of the ribaucour reduction of the fundamental transformation .
we finally remark that the equations characterizing the above lattices are relevant in physics , being integrable discretizations of equations arising in hydrodynamics and in quantum field theory .
integrable discrete geometry : the quadrilateral lattice , its transformations and reductions adam doliwa@xmath1 , paolo maria santini@xmath2 _ @xmath3instytut fizyki teoretycznej , uniwersytet warszawski + ul .
hoa 69 , 00 - 681 warszawa , poland _
@xmath4istituto nazionale di fisica nucleare , sezione di roma + p.le aldo moro 2 , i00185 roma , italy @xmath5dipartimento di fisica , universit di roma `` la sapienza '' + p.le aldo moro 2 , i00185 roma , italy @xmath6e - mail : [email protected] @xmath7e - mail : [email protected] |
the pierre auger observatory @xcite has led to great discoveries in the field of ultra - high energy cosmic rays ( uhecrs ) such as the confirmation of a suppression of the cosmic ray flux at energies above @xmath2 ev , limits on photon and neutrino fluxes at ultra - high energies and a hint of large scale anisotropies at energies above 8 eev .
nevertheless many questions related to these particles are still open . particularly interesting
is the behavior of the slant depth of the shower maximum with energy
. understood in terms of the lhc - tuned shower models , it suggests a gradual shift to a heavier composition , with a large fraction of protons at @xmath3 ev , changing to a heavier composition at @xmath4 ev @xcite .
however , we should interpret this result with caution since measurements of shower properties performed by the auger collaboration have revealed inconsistencies between data and present shower models .
for instance , the pierre auger collaboration has reported the first hybrid measurement of the average muon number in inclined air showers at ultra high energies , suggesting a muon deficit in simulations of about @xmath5 to @xmath6 at @xmath7 ev , depending on the hadronic interaction model @xcite .
hence the measured behavior of the slant depth of the shower maximum evolution could be understood as a hint of a new hadronic interaction physics at energy scales beyond the reach of lhc . in this work
we will deal with hadronic interactions in a statistical model , as first introduced by hagedorn @xcite ideas in the sixties . recently a power - law function based on the tsallis statistics @xcite has been widely used on fitting the transverse momentum ( @xmath8 ) and pseudo - rapidity ( @xmath9 ) distributions measured in high - energy collisions @xcite while several studies have been devoted to discuss these results in the literature @xcite .
the tsallis statistics , extensively used in different branches of science , is often used to describe systems which display properties like memory effects , long range interactions , intrinsic fluctuations , ( multi)fractal phase space and so on .
it consists in replacing the classical boltzmann - gibbs entropy @xmath10 by the form proposed by tsallis @xmath11 where @xmath12 is the probability of a particle occupy the state @xmath13 and @xmath14 is the tsallis index .
this definition comprises the boltzmann - gibbs entropy as a particular case , where @xmath15 . on the other hand , a straight consequence from this expression
is that the generalized entropy is no longer an extensive quantity , once we can verify that @xmath16 with the parameter @xmath14 being a measure of the nonextensivity of the system . as a consequence ,
we must replace the usual exponential boltzmann - gibbs distribution , @xmath17 , by the tsallis power - law distribution @xmath18^{\frac{1}{1-q}},\ ] ] where @xmath19 is the state energy and @xmath20 is the temperature of the system . according to @xcite
, the behaviors presented by the transverse momentum and pseudo - rapidity distributions , in high - energies domain , are best described using a nonexponential distribution , such as the one proposed by tsallis .
in fact , following the ideas discussed in @xcite , that behavior emerges from fluctuations of the thermal energy within the gas of quarks and gluons before the hadronization process . using this approach
, we can relate the parameter @xmath14 with those thermal fluctuations , @xmath21 obviously , when @xmath15 we recover the expected result obtained in the boltzmann - gibbs description , where we get an equilibrium at temperature @xmath20 . by assuming such scenario in which the temperature @xmath20 fluctuates within each collision
, the energy distribution of the particles generated in a single high energy interaction follows a power law tsallis distribution , given by eq .
the left panel of figure [ histo ] presents the tsallis energy distribution @xmath22 with a fixed temperature @xmath20 for different values of @xmath14 .
we can see that higher the @xmath14 values , the probability for generating particles with larger energy values become greater . as a consequence of the total energy conservation constraint , @xmath23 , where @xmath24 is the total energy of the interaction in the center of momentum frame
, it can be shown that the tsallis statistics leads to a negative binomial multiplicity distribution given by @xmath25^{-n + \frac{1}{1-q}}.\ ] ] such distribution has a form shown in the right panel of the figure [ histo ] , where it is possible to see how its maximum is affected by @xmath14 , becoming closer to zero as @xmath14 grows . besides , the inset plot of this figure shows that the relationship between the value for the maximum of the multiplicity distribution and @xmath14 is quite linear , at least in that domain of @xmath14 values .
therefore , one can see that the introduction of the tsallis statistics in this context changes the energy , momenta and multiplicity distributions of the particles generated in the hadronic interaction
. for a fixed value of temperature @xmath20.,title="fig : " ] for a fixed value of temperature @xmath20.,title="fig : " ] the transverse momentum and pseudo - rapidity distributions resulting from high energy collisions measured by several experiments show a large discrepancy in the values of the parameter @xmath20 , reflecting different physics for the transverse and the longitudinal space .
the transverse distributions are thermal - like , presenting a parameter @xmath26 almost independent of the energy while those from the longitudinal space have a temperature sensitive to the energy of the collision , understood as the mean energy available per produced particle @xcite , @xmath27 , where @xmath28 and @xmath29 are , respectivelly , the mean multiplicity and inelasticity of the interaction . moreover , since the measured tsallis index for the longitudinal space @xmath30 is much larger than measured for the transverse space @xmath31 and @xmath32 , the resulting @xmath33 .
also , as verified by simulations , the transverse momentum distribution has a minor contribution on the cosmic ray observables studied in this work .
therefore , from now on , we will assume a statistical equilibrium for the transverse momentum space and we will refer to the entropic index @xmath34 and temperature @xmath35 .
the goal of the present paper is to study the impact of the temperature @xmath20 fluctuations , represented by the parameter @xmath14 , on the shower maximum , @xmath0 , and number of muons on the ground , @xmath1 .
the simulations performed in this work are described in section [ simulations ] . in section [ results ]
we present the results of the simulations and discuss them in light of the heitler model .
finally , we present the conclusions of this work in section [ conclusions ] .
for all simulations presented in this work , we have used corsika 7.40 @xcite with the interaction models sibyll 2.1 @xcite and gheisha 2002d @xcite , for high and low energy processes , respectively .
the muon energy threshold used in the simulations is 0.3 gev and the array detector position is at 1400 m above sea level , corresponding to the mean altitude of the pierre auger observatory .
the air shower simulation chain is as follows : first we simulate the secondaries generated in the collision between a cosmic ray and a nucleus of the upper atmosphere externally by assuming that the hadronization process is described by the tsallis statistics ; the resulting particle list is then inserted back into corsika ( using the stacking option and sampling option with thinning = @xmath36 ) to proceed with usual cascade development through the atmosphere .
such a procedure was performed 1000 times for each of several values of @xmath14 ( 1.01 , 1.025 , 1.05 , 1.075 , 1.10 , 1.15 , 1.20 , 1.25 , 1.30 , 1.35 , 1.40 , 1.45 and 1.50 ) for showers with zenith angle @xmath37 initiated by a proton of fixed energy @xmath38 ev .
the reason for limiting the entropic index to @xmath39 in this work is that the mean value of the tsallis distribution @xmath22 , given by @xmath40 , is well defined only for @xmath41 @xcite .
this model assumes that the sibyll predictions are valid for lower energies since they are tuned by accelerator data while it fails for higher energies .
this added to the parametrizations of the lhc transverse momentum and pseudo - rapidity distributions by the tsallis statistics justifies its use in this work for energies above @xmath42 ev .
the point of the air shower first interaction is determined using the @xmath43-air cross - section predicted by the epos 1.99 model @xcite .
the reason for using this value instead of the one predicted by the sibyll model is that the later presents a large discrepancy in relation to that measured by the pierre auger collaboration @xcite .
the mean multiplicity @xmath28 and inelasticity distribution of the p - air interaction used in this work were extracted from p - air interaction simulations using the sibyll model .
since the tsallis distribution is non - extensive , generating particle energies @xmath44 according to this distribution subject to the constraint @xmath23 is not a simple task because the probabilities @xmath45 associated to each particle do not factorize @xcite .
therefore , we perform the simulation process according the following procedure : first , we select the number of particles generated in the p - air interaction using the @xmath46 expression given by eq .
[ p_n ] and we assign an energy @xmath44 to each particle @xmath13 according to the tsallis distribution .
then two particles @xmath13 and @xmath47 are randomly selected and a random fraction @xmath48 of the energy @xmath44 is given to particle @xmath47 in such a way that the new values of energies are @xmath49 and @xmath50 .
after that , we compute the deviation of this energy distribution in relation to tsallis distribution @xmath22 using the @xmath51 estimator defined by : @xmath52 if the new @xmath51 value is smaller than the previous one , we accept the changes in energy of the particles @xmath13 and @xmath47 , otherwise we cancel them .
we keep repeating this procedure for another pair of particles .
the whole process continues until the @xmath51 value is stabilized .
generally , it takes @xmath53 iterations to reach such stabilization . to be conservative
, the simulations presented in this work were performed using @xmath54 iterations for each p - air interaction . since we verified through simulations that @xmath0 and @xmath1
are not sensitive to changes in @xmath31 , all simulations corresponding to the transverse space presented in this work were evaluated assuming a statistical equilibrium , i.e , the hagedorn @xcite transverse momentum distribution : @xmath55 with @xmath56 mev.the type of particles are randomly generated according to sibyll predictions and once we have generated the particles masses @xmath57 , the longitudinal momentum is obtained as @xmath58 .
these kinematic variables along with the species of particle complete all the information we need to reintroduce in corsika to proceed with shower propagation through the atmosphere .
in the following we describe the impact of @xmath59 fluctuations , represented by the parameter @xmath34 , on the shower maximum , @xmath0 , and number of muons on the ground .
we will discuss them in terms of the predictions of the heitler model @xcite as well as of the results achieved in reference @xcite .
although extremely simple , the predictions of the heitler model are remarkable .
it assumes that the shower maximum is reached when the energies of particles become smaller than a critical energy , in which energy loss processes dominate the production of new particles in the case of electromagnetic component , or the charged pion interaction length becomes larger than the decay length of pions in muons , in the case of the hadronic one . as a consequence , it predicts an increase of the @xmath60 for smaller mean multiplicities , since larger multiplicities correspond to lower energy per particle . besides , reference @xcite describes a detailed investigation of the impact of the multiplicity , hadronic particle production cross section , elasticity and pion charge - ratio on air shower observables with most of the predictions qualitatively understood within the simple heitler model and its extension to hadronic component .
the pearson coefficient @xmath61 was used to assess the degree of correlation between air showers observables and @xmath14 .
it is defined by @xmath62 and measures the linear correlation between two variables @xmath63 and @xmath64 , yielding a value in the interval @xmath65 $ ] , with @xmath66 meaning total positive correlation , @xmath67 no correlation , and @xmath68 total negative correlation .
cov(x ) is the covariance between @xmath63 and @xmath64 and @xmath69 and @xmath70 are the standard deviations of variables @xmath63 and @xmath64 .
the results for the mean depth of shower maximum , @xmath60 , and the fluctuations of @xmath0 are summarized in figure [ dist_xmax ] .
the left panel shows the @xmath60 as a function of @xmath14 .
a strong correlation is observed yielding a pearson correlation coefficient @xmath71 .
the red line presents the best linear fit corresponding to a @xmath72 .
the comparison between the predictions from heitler model and reference @xcite with our results requires caution since we did not change the mean multiplicity in our simulations of the first interaction .
however , the changes in distributions of energy and momenta of the particles generated in the first interaction results in a spread of the multiplicity distribution and a shift of its peak to lower values as it is shown in figure [ histo ] .
the reason for the strong correlation is that most of the showers generated with larger @xmath14 values are initiated with smaller multiplicities , or equivalently , with higher energy per particle .
besides , the right panel of figure [ dist_xmax ] presents the corresponding plot for the @xmath73 as a function of @xmath14 . in this case , the observed correlation is weak , with @xmath74 and @xmath75 corresponding to the best linear fit , shown by the red line . according to the heitler model , the variance of the @xmath76 distribution depends on the hadronic cross - section interaction , @xmath77 , and multiplicity , @xmath78 , via @xmath79 , where @xmath80 g/@xmath81 is the electromagnetic radiation lenght . in this work ,
the hadronic interaction is given by the one corresponding to the interaction between a proton and a nucleus of the atmosphere , @xmath82 .
although the @xmath83 increases with @xmath14 , the observed correlation between @xmath84 and @xmath14 is weak since the increase of spread of the multiplicity distribution for larger @xmath14 values is dominated by the first term contribution in @xmath84 , as a consequence of the relatively small value of the p - air cross - section .
the impact of @xmath14 on the mean number of muons in the ground , @xmath85 , and the fluctuations of @xmath1 are summarized in figure [ dist_mu ] .
the @xmath85 as a function of @xmath14 , shown in the left panel , presents a strong anti - correlation with @xmath14 , with @xmath86 and @xmath87 corresponding to the best linear fit , marked in red .
a superficial analysis of the left panels of figures [ dist_xmax ] and [ dist_mu ] could indicate wrongly that @xmath85 and @xmath60 are anti - correlated .
the positive correlation between @xmath85 and @xmath60 position exists but it is weak since muons are hardly attenuated in the atmosphere .
therefore it is not the most important factor for the @xmath85 behavior as a function of @xmath14 .
indeed muons are mainly produced as a result of pion decay and their abundance in the ground , especially considering the most energetic particles , is strongly correlated with the number of pions in the shower . as a consequence of the reduction of the peak of the multiplicity distribution for larger @xmath14 values ,
most showers presents lower production of pions in the first interaction , constituting the main reason for the observed anti - correlation between @xmath85 and @xmath14 . on the other hand
, the right panel of figure [ dist_mu ] shows a strong correlation of @xmath88 and @xmath14 , with @xmath89 , as a natural consequence of the spread of the multiplicity distribution .
the red line shows the best linear fit with corresponding @xmath90 .
finally , table [ summary_results ] summarizes the correlation coefficients @xmath61 between air shower observables and @xmath14 obtained in this work .
.summary of the correlation coefficients between air shower observables and @xmath14 obtained in this work . [
cols="^,^,^,^,^",options="header " , ]
although the simulations presented in this work are a very simple description of ultra high energy interactions , the results presented here show that intrinsic fluctuations of the system with respect to @xmath59 , given by the parameter @xmath34 , change the energy , momenta and multiplicity distributions of the particles generated in the interaction between a cosmic ray and a nucleus of the atmosphere , impacting air shower observables such as the slant depth of the maximum @xmath0 and the muon number on the ground @xmath1 .
the results show that the higher the temperature fluctuations , the greater the values of the mean slant depth of maximum @xmath91 and variance of the number of muons on the ground @xmath92 , with pearson correlation coefficients of @xmath71 and @xmath89 , respectively .
this results from the spread and shift of the maximum of the multiplicity distribution to lower values for larger temperature fluctuations .
besides , as muons are mainly produced by the decay of charged pions and the shift in the peak of multiplicity distribution reduces the number of such particles generated in the first interaction , the mean number of muons on the ground @xmath93 presents a strong negative correlation with @xmath14 , producing a @xmath86 . on the other hand ,
the variance of the slant depth distribution @xmath94 presents a very weak correlation with the temperature fluctuations , with @xmath74 , because the contribution of the hadronic p - air interaction fluctuations dominates the one originated from the multiplicity distribution .
these results agree qualitatively with the heitler model and reference @xcite predictions .
although these results have been obtained for a specific non - extensive hadronic interaction model , we believe that it captures the essential features related to the presence of tsallis statistics in uhecr showers and can shed light to the understanding of their properties as well as of particle interactions at these energies .
studies regarding different non - extensive particle interaction models are out of the scope of this work and will be addressed in a future work .
we thank professor j. c. anjos for useful discussions in the beginning of this work . this work was partially supported by the conselho nacional de desenvolvimento cientfico e tecnolgico " ( cnpq ) and by the
fundao carlos chagas filho de amparo pesquisa do estado do rio de janeiro " ( faperj ) .
pierre auger collaboration , j. abraham et al . , _ properties and performance of the prototype instrument for the pierre auger observatory " _ , nucl .
instrum . meth . a 523 , 50 ( 2004 ) pierre auger collaboration , a. aab et al . , _ the pierre auger cosmic ray observatory , " _ nucl .
instrum . meth . a 798 172 ( 2015 ) pierre auger collaboration , a. aab et al .
`` muons in air showers at the pierre auger observatory : mean number in highly inclined events '' _ , phys . rev .
d * 91 * ( 2015 ) 032003 ; erratum : phys .
d * 91 * , 059901 ( 2015 ) .
j. cleymans , g. i. lykasov , a. s. parvan , a. s. sorin , o. v. teryaev and d. worku,_``systematic properties of the tsallis distribution : energy dependence of parameters in high - energy p - p collisions . ''
b * 723 * , 351 ( 2013 ) [ arxiv:1302.1970 [ hep - ph ] ] .
j. cleymans , _ `` the tsallis distribution for p - p collisions at the lhc '' _ , j. phys .
455 * , 012049 ( 2013 ) ; m. d. azmi and j. cleymans , _ `` transverse momentum distributions in p - pb collisions and tsallis thermodynamics '' _ , arxiv:1311.2909 [ hep - ph ] ; m. d. azmi and j. cleymans , _
`` transverse momentum distributions in proton - proton collisions at lhc energies and tsallis thermodynamics '' _ , arxiv:1401.4835 [ hep - ph ] .
g. wilk and z. wlodarczyk , _
`` interpretation of the nonextensivity parameter q in some applications of tsallis statistics and lvy distributions '' _ , phys .
lett . * 84 * , 2770 ( 2000 ) [ hep - ph/9908459 ] .
l. marques , j. cleymans , a. deppman , _ `` description of high - energy pp collisions using tsallis thermodynamics : transverse momentum and rapidity distributions '' _ , phys.rev .
d * 91 * , 054025 ( 2015 ) [ arxiv:1501.00953 [ hep - ph ] ] .
bao - chun li , ya - zhou wang , fu - hu liu , xin - jian wen , you - er dong , _ `` particle production in relativistic @xmath102(@xmath103 ) and aa collisions at rhic and lhc energies with tsallis statistics using the two - cylindrical multisource thermal model '' _ , phys.rev .
d * 89 * , 054014 ( 2014 ) [ arxiv:1403.4025 [ hep - ph ] ] .
bao - chun li , zhao zhang , jun - hui kang , guo - xing zhang , fu - hu liu , _ `` tsallis statistical interpretation of transverse momentum spectra in high - energy pa collisions '' _ , adv.high energy phys . * 2015 * , 741816 ( 2015 ) .
k. werner , f. m. liu , and t. pierog , _
`` parton ladder splitting and rapidity dependence of transverse momentum spectra in deuteron - gold collisions at the bnl relativistic heavy ion collider '' _ , phys .
c * 74 * , 044902 ( 2006 ) . | we investigate the impact of the tsallis non extensive statistics introduced by intrinsic temperature fluctuations in p - air ultra high energy interactions on observables of cosmic ray showers , such as the slant depth of the maximum @xmath0 and the muon number on the ground @xmath1 .
the results show that these observables are significantly affected by temperature fluctuations and agree qualitatively with the heitler model predictions . |
we study topological chaos in regimes where two - dimensional flow is driven by moving stirrers , such as in a prototypical batch stirring device @xcite illustrated in figure [ fig : mixer ] .
we consider the effect of extra stirrer motions that are possible when the fluid domain has spatial periodicity .
for the mixer in figure [ fig : mixer ] , we imagine removing one or both pairs of opposite walls and identifying these edges to make a periodic domain .
our study is motivated by the many model flows examined in the literature that have spatial periodicity , such as the ubiquitous sine flow @xcite .
some laboratory experiments are also designed to have spatial periodicity in one direction @xcite , and the square patterns formed in thermal convection exhibit periodicity in two directions @xcite .
there are also signal analysis applications of braid groups @xcite where measured quantities can be periodic .
topological chaos ( tc ) can be used to design mixers with a robust built - in stirring quality that depends only on the relative motions of the stirrers and is independent of fluid properties such as compressibility or viscosity @xcite .
this mixing quality can be guaranteed because stirrers are topological obstacles to flow : if their trajectories form an intertwined braiding pattern , then by continuity material lines in the fluid must also be braided .
if the stirrer motion traces a braid with positive topological entropy , then typical material lines in the mixing region of the fluid will grow exponentially . roughly speaking
, this means that the braiding motion of the stirrers guarantees a minimum amount of chaos in the flow in a region around the stirrers , and chaos leads to efficient fluid mixing @xcite . in the two - dimensional flows studied here ,
the topological entropy is related to the time - asymptotic exponential growth rate of material lines .
since the length of material lines increases rapidly , they must fold back and forth within the bounded mixing region and lead to the familiar `` striations '' observed in typical mixing flows .
the properties of a braid are linked to thurston
nielsen theory @xcite .
thurston proved that a two - dimensional homeomorphism ( here , the map giving the motion of fluid particles in one period ) can be decomposed into regions that are either reducible , finite - order , or isotopic to a pseudo - anosov map .
braids are a convenient language to describe the _
isotopy classes _ of these homeomorphisms . for our purposes ,
the third type is most important : pseudo - anosov maps have strong chaotic properties and typically lead to good mixing . if the braid is _ irreducible _ with positive entropy ,
then to it must correspond a region of the flow that is a single chaotic component . in fluid mixing parlance , this region will be well - mixed ( though it could be very small or even have zero measure ) . to guarantee positive topological entropy in a nonperiodic domain , it can be shown that at least three trajectories must be involved in a braid . however , this does not imply that mixers must have three or more stirrers to produce tc @xcite .
this is because we can consider trajectories of anything that acts as a topological obstacle to material lines : in a two - dimensional flow this could be a fluid particle itself , or an entire island of regularity .
thus , all mixers generally produce tc , since it is easy to find sets of periodic trajectories that braid .
this means that topological ideas can also be used to understand mixing , even for flows with fewer than three stirrers @xcite .
it is also instructive to note that all chaos in time - periodic two - dimensional flows is topological , in the sense that the topological entropy of a flow can be arbitrarily well - approximated by considering the braiding motion of a sufficiently high - order periodic orbit @xcite .
we begin in section [ sec : bounded ] by summarising current techniques that have been applied to bounded flows ( or spatially - nonperiodic flows on the infinite plane ) .
then in section [ sec : periodic ] we describe ways of extending these methods for spatially periodic flows , examining in turn the case of single- and double - periodicity . in section [ sec : numerics ] we determine the sharpness of theoretical topological entropy estimates compared to entropies for model stokes flows .
we apply our techniques to the sine flow in section [ sec : sineflow ] , where the `` stirrers '' are given by periodic orbits .
finally , we close with a discussion of our results in section [ sec : discussion ] .
in this section we describe how motion of stirrers around each other in a nonperiodic domain ( _ i.e. _ , a surface of genus zero such as a bounded domain or the infinite plane ) may be described as a braid .
we then review how this braid may be used to produce a rigorous lower bound on the topological entropy of the flow .
we consider two - dimensional flows , where the topological entropy is the time - asymptotic growth rate of material lines @xcite .
( it is the supremum over all possible smooth material lines , but in practice we find that the growth rate rapidly becomes independent of the choice of initial material line , except for pathologically bad choices . ) in flows without spatial periodicity , braiding is conventionally characterised by recording exchanges of position of adjacent stirrers @xcite .
stirrer positions are projected onto an axis ( we have chosen the @xmath0-axis in what follows ) and ordered from @xmath1 to @xmath2 along this line . if stirrer @xmath3 and @xmath4 exchange order in a clockwise direction , we assign to this motion the braid group generator @xmath5 ; an anti - clockwise crossing is labelled @xmath6 .
the numbers @xmath1 to @xmath2 always refer to the relative position of the stirrers along the @xmath0-axis , and do not always label the same stirrer .
the procedure of associating trajectories with braid generators is explained in detail in refs .
any motion of stirrers then corresponds to an ordered sequence of generators : this sequence is an element of artin s braid group @xcite on @xmath2 strands .
this group contains all braids generated by sequences of @xmath7 and their inverses .
the group operation is the catenation of braids , and the inverse is found by reversing the sequence and inverting the generators .
the identity of the braid group is simply @xmath2 parallel strands .
we use the convention that braid generators occur from left to right , so that in the braid word @xmath8 the generator @xmath9 occurs first .
some braids are topologically equivalent because they can be continuously deformed into each other .
in fact , equivalence of braids can be established using just the group relations @xmath10 which together form a presentation for the group ( a minimal set of relations describing the group ) .
relation implies that particular sequences of three braid elements are equivalent , and is obvious by inspecting a diagram of the braid @xcite .
relation dictates that any two braid events involving different pairs of stirrers can be commuted .
the braid word describing the relative motions of the stirrers encodes a great deal of information about the fluid flow .
one such piece of information is the topological entropy of the braid , which is a lower bound on the topological entropy of the flow .
two methods of calculating this lower bound have been considered : matrix representations of the braid group @xcite , and train - tracks @xcite . for three stirrers ,
the two are equivalent ; for more than three stirrers the matrix method only gives a lower bound on the topological entropy of the braid .
hence in this paper we shall use exclusively train - tracks to compute the topological entropy of a braid .
this is discussed in the following sections .
a complicated but robust method for computing braid topological entropies is the bestvina
handel train - track algorithm @xcite .
this algorithm requires the construction of a special invariant graph for each braid .
computerised implementations of this algorithm exist that take a word in generators of artin s braid group as an input @xcite .
these are not currently well - suited to the periodic domains that we consider here , though they could be adapted for this purpose . here
we shall be dealing with braids that are simple enough that we will not need the bestvina
handel algorithm .
rather , we will use the method that thurston calls ` iterate and guess , ' which is based on the fact that under iteration almost any curve converges to the train - track ( for pseudo - anosov braids ) .
a train - track consists of a set of edges , joined together at junctions . for the braids studied in this paper the following types of edge
are needed : 1 .
edges that wrap tightly around the stirrers in a closed loop starting and finishing at the same junction ; 2 .
edges that join the junctions associated with different stirrers .
the edges of a train - track graph are deformed under the action of the braid , and we assume the edges are like elastic bands , so that they stretch by the smallest amount possible without crossing through any of the stirrers .
type 1 edges , since they are wrapped tightly around the stirrers , are not stretched under the action of the braid . however , under the action of the braid the stirrers may be permuted , so type 1 edges are permuted .
type 2 edges are non - trivially deformed under the braid and may become wrapped around several stirrers .
a graph is said to be invariant if under the action of the braid all edges are mapped onto edge - paths of the original graph .
in general , type 2 edges are mapped onto a edge - paths made up of type 1 and type 2 edges . under the action of the flow , a material line representing edges from the train - track
must be stretched at least as much as the train - track itself .
however , in general it is stretched even more due to the presence of additional periodic orbits which act as obstacles to the flow ( more on this in section [ sec : discussion ] ) .
to summarise the technique so far : a graph of type 1 and 2 edges is drawn between and around the stirrers .
then the stirrers are made to undergo the braid operation under investigation , and the deformation of the graph is tracked as if its edges were rubber bands .
the goal is to find a graph that is invariant under the action of the braid that is , the edges may stretch and fold , but they all end up at positions where there was initially an edge - path .
of course , the edges get permuted and stretched , so a single edge may be mapped to a path that traverses the original edges several times .
we call the invariant graph a train - track for the braid .
( in order to be a genuine train - track , a graph must have a special non - cancellation property that we discuss below . )
once the train - track graph has been found , the stretching in the train - track may be measured by constructing a transition matrix describing for each edge the edge - path into which it is mapped .
the logarithm of the largest eigenvalue of this matrix is the topological entropy of the braid .
the largest eigenvalue is used because it is assumed that the braid operation is repeated a large number of times , so the graph converges to an invariant pattern corresponding to the eigenvector of the largest eigenvalue .
the entries in the eigenvector describe the relative number of edges that accumulate under repeated action of the braid .
the eigenvalue then describes the growth of the length of this invariant pattern at each application of the braid . as an example we consider the well - studied pigtail braid @xmath11 with three stirrers , with train - track illustrated in figure [ fig : btrack](a ) .
it consists of three type 1 edges , labelled 1 , 2 and 3 , that pass around each of the stirrers , and two type 2 edges , labelled @xmath12 and @xmath13 , that join adjacent stirrers .
the two sequences shown in figure [ fig : btrack](b ) illustrate the fate of the edges @xmath12 and @xmath13 under @xmath9 followed by @xmath14 .
up to a deformation that does not change the topology , the edges of the train track are mapped to the following edge - paths : @xmath15 an important feature to note in this example is that ultimately the edges @xmath12 and @xmath13 stretch from the first stirrer , _ underneath _ the second one , to the third one ; hence the resultant edge - paths include a tour around edge @xmath16 . in every train - track all edges must be mapped to an edge - path that has alternately type 2 edges ( denoted by letters ) and type 1 edges ( denoted by numbers ) .
this ensures that two or more adjoining type 2 edges do not collapse back to a shorter edge - path under repeated action of the braid .
( this is the non - cancellation property alluded to above . ) with no possibility of such cancellations a transition matrix can be written down by recording the number of edges in each edge - path .
this matrix can then be used to determine the evolution of an arbitrary edge - path under any number of subsequent applications of the braid .
the transition matrix for the braid word @xmath17 is found from to be @xmath18
, \label{eq : transmx}\]]where the element @xmath19 represents the number of occurences of the @xmath3th edge in the image of the @xmath20th edge , and the indices @xmath3 and @xmath20 take values @xmath21 , in that order .
let the vector @xmath22 contain the number of edges of each type in a path . under the action of the braid
these will be mapped to a longer path with edges given by @xmath23 .
the topological entropy of the flow is given in two dimensions by the time - asymptotic growth rate of material lines ( maximised over all possible smooth lines ) .
since the train - track could be considered as a material line , the growth rate of the flow must be at least as large as the growth rate of the train - track .
this is given by the spectral radius ( the magnitude of the largest eigenvalue ) of the transition matrix , because for a large number of repeated applications of the braid the coiling will become dominated by the largest eigenvector of the transition matrix .
for the pigtail braid , the matrix has a spectral radius of @xmath24 .
hence the topological entropy of the braid @xmath17 is .
any time - periodic flow which has material obstacles executing this braiding motion necessarily has a topological entropy of at least @xmath25 .
note that since type 1 edges are wrapped tightly around stirrers , and are only permuted amongst themselves under the braid , they do not contribute to growth of material lines .
it is thus usual to write the transition matrix only for type 2 edges . in the present example
this is the upper - left block indicated in .
in this section we study the braid group on the cylinder and the torus . because of the possibility of stirrers going around the periodic direction(s ) , these groups are different to artin s braid group discussed in section [ sec : bounded ] .
note that thurston
nielsen theory has been considered before on the annulus and torus , see for example @xcite and @xcite . here
we focus on a braid description of stirrer motion on these surfaces , using for the most part the notation of birman @xcite . for a periodic domain , in addition to the braid elements @xmath5
representing the exchange of stirrer positions , we introduce extra motions corresponding to stirrers making tours of the periodic directions , as illustrated in figure [ fig : zplane ] . in a singly - periodic domain
we allow the extra braid motions @xmath26 corresponding to the @xmath3th stirrer touring the periodic direction .
the singly - periodic case represents flow on a ( possibly infinite ) cylinder . in this paper
we choose the periodicity of the cylinder to be in the @xmath0-direction without loss of generality . for a doubly - periodic flow
we allow the extra braid motions @xmath26 and @xmath27the additional motions @xmath27 correspond to the @xmath3th stirrer touring the second periodic direction .
the doubly - periodic case can be thought of as flow on a torus .
the braiding motions @xmath26 and @xmath27 are not independent of the @xmath5 nor each other , since some braids can be continuously deformed into each other .
the braid group relations for an arbitrary number of stirrers have been stated by birman @xcite . as an example , for the braid group on the torus with two stirrers , the one relation required to determine equivalence of braids is @xcite @xmath28 the relation provides a presentation for the two - strand braid group on the torus , with generators @xmath29 .
the other two operations can be written @xmath30 , @xmath31 .
the ideas developed in the following section apply for an arbitrary number of stirrers , but we shall illustrate flow with just two stirrers .
we will see that is possible to create topological chaos on a periodic domain with fewer than three stirrers .
for the special case of the cylinder , we consider braids formed from @xmath5 and @xmath26 generators only , and ignore @xmath27 for the moment .
the dynamics on a singly - periodic domain is only a little more complicated than the dynamics in a bounded domain .
this is because , as pointed out in ref .
@xcite , there exists a conformal mapping from the periodic strip ( _ i.e. _ , the cylinder ) to an annulus , and the topological entropy of a flow is preserved under this conformal mapping . in figure
[ fig : wplane ] we show the result of conformally mapping the square in the @xmath32-plane ( with @xmath33 ) in figure [ fig : zplane ] under the mapping @xmath34 onto an annulus in the @xmath35-plane .
the stirrers aligned along a diagonal are mapped onto a logarithmic spiral in the @xmath35-domain .
the key feature in this @xmath35-domain is that there is a hole ( centered on the origin ) that acts as an @xmath36th stirrer . in the case of the @xmath32-domain extending to @xmath37 this hole shrinks to a point , but it is still a topological obstacle to the flow .
if the @xmath32-domain includes @xmath38 then the @xmath35-domain extends to infinity .
braid elements @xmath5 do not involve the hole .
however , the periodic motions @xmath26 make a clockwise tour of the hole ( stirrer @xmath39 ) , also crossing all stirrers @xmath40 twice ( figure [ fig : wplane ] ) . thus the braid group on a singly - periodic domain with @xmath2 strings , with generators @xmath41 , can be associated with the braid group in the plane with @xmath39 strings , with generators @xmath42 , if we assign @xmath43 an important consequence of the hole acting as a stationary stirrer is that only two proper stirrers are required to guarantee topological chaos .
( on a nonperiodic domain , at least three stirrers are needed , since the topological entropy of all braids on two strands is zero @xcite . )
we illustrate this for the two - stirrer braid @xmath44 which is equivalent to the three - stirrer planar braid @xmath45 .
we will show in the following section that this braid has a topological entropy of @xmath46 .
note that this is larger than the entropy @xmath25 for the same length braid word @xmath17 , which was shown in ref .
@xcite to be optimal on a nonperiodic domain .
this suggests that periodic boundary conditions contribute significantly in creating chaos ! in principle we can compute the topological entropy of braids on the cylinder by first applying the conformal transformation defined in section [ sec : cylinder ] and finding the corresponding braid word on the annulus .
however , the braid word on the annulus is usually considerably longer than that on the cylinder , which makes the train - track harder to compute .
moreover , for the doubly - periodic case ( torus ) there is no such conformal mapping trick , so in any case we will need to find train - tracks on periodic domains .
fortunately we can find train - tracks directly for the cylinder and torus by considering graphs with edges that traverse the periodic directions .
the periodic train - tracks shown in this paper are determined by inspection , following thurston s ` iterate and guess ' approach , though we expect that existing computer codes ( see _ e.g. _
@xcite ) could be extended to automate their construction .
the method works as follows .
consider stretching an elastic band between two stirrers , possibly traversing the periodic direction in doing so . now consider the fate of the elastic under the action of the braid .
the band will be stretched around the stirrers with the same topology as the train - track for the braid ( except for possible rare choices of initial condition where the band undergoes a periodic motion without stretching at all ) . if an invariant train - track exists then the elastic band will tend to align with it as it is stretched .
the reason it must align is that initially the elastic crosses the train - track a fixed number of times ( possibly zero ) , and by determinism it must still cross the same number of times even after many applications of the braid . now ,
if the elastic is rapidly stretching but never makes any additional crossings of the train - track , then it must align and be stretched parallel to the track .
after a few iterations of the braid it is usually possible to determine the shape of the required invariant graph by inspection .
it can then be verified that the graph is a proper train - track by checking that all type 2 edges are mapped onto edge - paths consisting of alternating type 2 and type 1 edges . in figure
[ fig : strack ] we illustrate construction of a train - track for the braid @xmath44 studied above .
the starting point for this calculation is to compute the fate of an elastic band @xmath12 stretched directly between the two stirrers . under the action of the braid this is deformed into a ` z ' shape , shown in the top sequence in figure [ fig : strack](b ) .
note that headaches can be spared by drawing extra copies of the periodic domain as necessary , rather than attempting to wrap the line around on a single copy of the domain .
the resulting ` z ' shape has two edges like the original @xmath12 , and a new edge call it @xmath13joining stirrer @xmath1 to @xmath16 across the periodic direction . the bottom sequence in figure [ fig : strack](b ) shows the evolution of @xmath13 .
the edge @xmath13 is mapped to a double ` z ' shape that spans three copies of the domain and consists of three @xmath12 edges and two @xmath13 edges . because no unfamiliar edges are generated we conjecture that @xmath12 and @xmath13 , plus two type 1 loops to allow lines to wrap around the stirrers , are all that is required for the train - track . to confirm that this invariant graph is a proper train - track we examine the edge - paths produced by the braid , which are @xmath47 all the type 2 edges
are mapped to paths of alternating type 2 and type 1 edges , so this confirms we have found the required train - track .
the transition matrix for the braid @xmath44 ( ignoring type 1 edges which do not contribute to stretching ) is then @xmath48\ ] ] with spectral radius @xmath49 , so the braid has a topological entropy of @xmath46 .
the same inspection method works for toroidal braids , where we allow the additional braid elements @xmath27 corresponding to stirrers touring the second periodic direction .
we illustrate the construction in figures [ fig : dtrackab][fig : dtrackedge ] for the more complicated braid @xmath50 .
figure [ fig : dtrackab ] depicts the invariant train - track for this braid .
the action of the braid on the various edges is shown in figure [ fig : dtrackedge ] .
our initial ` guess ' elastic band is labelled @xmath12 . under the action of the braid ,
@xmath12 is mapped onto three new edges , @xmath13 , @xmath51 , and @xmath52 ( figure [ fig : dtrackedge ] , top ) .
edge @xmath13 is mapped to a path containing @xmath13 , @xmath51 and @xmath52 .
edge @xmath51 is mapped onto @xmath13 and @xmath52 .
edge @xmath52 is mapped onto @xmath12 , @xmath13 and @xmath52 , and no unfamiliar edges are generated .
hence @xmath12 , @xmath13 , @xmath51 and @xmath52 , plus two small loops around the stirrers , may be checked as a candidate train - track .
careful inspection reveals edges are mapped to edge - paths as @xmath53 each of these is of the required form ( that is , type 1 edges alternate with type 2 ) , so this is the required train - track .
the corresponding transition matrix ( keeping only the finite - length type 2 edges ) for @xmath54 is then @xmath55\ ] ] with spectral radius @xmath56 , so the braid has a topological entropy of @xmath57 .
this growth rate is slightly greater than that achieved by performing the cylinder braid @xmath58 twice ( resulting topological entropy @xmath59 ) , at the same energetic cost ( by symmetry ) .
however , no attempt has been made at optimising this torus braid , so braids with higher topological entropy may exist .
in this section we show numerical simulations of quasi - steady two - dimensional spatially periodic stokes flow to determine the sharpness of predicted entropy lower bounds compared to actual flow topological entropies . before investigating a particular flow regime ,
it is important to remember that the topological entropy bounds given above are based only on the braiding motion of the stirrers and are independent of all properties of the underlying fluid flow except for continuity . in a given flow regime , however , we expect that there could be additional features that lead to more intricate braiding , and therefore to an improved topological entropy . in our simulations , entropies are computed numerically by recording the stretched length @xmath60 of a material line as a function of time @xmath61 . because material lines are stretched exponentially fast in a chaotic flow , the slope of a best - fit line through @xmath62 versus @xmath61 gives the topological entropy of the flow ( after ignoring initial transients ) .
we consider a very viscous ( stokes ) fluid in a unit square domain @xmath63 .
stirring is performed with just two stirrers , since we have shown above that this is sufficient to allow topological chaos in a periodic domain .
the stirrers are circular with diameter @xmath64 , and are initially centred at @xmath65 and @xmath66 . in the stirring operations @xmath9 , @xmath67 and @xmath68 the stirrer trajectories we prescribe are given in table [ tab : rodtraj ] .
note that in the stokes flow regime the resultant advection depends only upon the paths of the stirrers and not on the stirrer velocities .
.trajectories of the stirrers in the simulations corresponding to the braid generators , with and . [ cols="<,^,^,^,^",options="header " , ] [ tab : rodtraj ] the stokes flow velocity field was computed using the lattice - boltzmann method @xcite . whilst a lattice - boltzmann code is relatively inefficient for computing stokes flow
, it copes easily with the moving circular boundaries and also with periodic boundary conditions , in addition to allowing investigation of similar flows at moderate reynolds number ( not reported here ) .
we have chosen to compute the velocity field this way because the complex variable series solution approach employed by finn _ et al . _
@xcite does not generalise easily for periodic domains . in computing the velocity field , for simplicity
we impose a spatially periodic pressure field , and thus find the unique solution that has no net flux in either of the periodic directions .
we consider passive advection of an initial closed polygon with corners at @xmath69 , @xmath70 , @xmath71 , @xmath72 , @xmath73 and @xmath74 , with a total length of @xmath75 .
this represents the interface between a two - by - two checkerboard pattern of black and white squares of fluid .
as the interface is advected we use a particle insertion scheme to ensure that stretches and folds remain well resolved @xcite . in figure
[ fig : singly ] we show the result of stirring using the braid @xmath58 in a domain that is periodic in the @xmath0-direction , with no - slip boundaries at @xmath76 .
figure [ fig : doubly ] shows the evolution of the fluid under the braid @xmath50 in a doubly - periodic domain . in both figures
it is clear that the fluid is well mixed .
closer inspection of the snapshots after complete application of each braid also reveals accumulations of thin striations that roughly align with the train - track graph .
this is more obvious for the singly - periodic case shown in figure [ fig : singly ] , where the thinnest striations form in a ` z ' shape similar to that shown in the corresponding train - track in figure [ fig : strack ] .
it was noted by alvarez _
_ @xcite that in chaotic flows striations accumulate with different densities in different regions of the domain . here , in the long - time limit
the relative number of striations caught between stirrers is given by the entries of the dominant eigenvector of the train - track transition matrix .
this relative number is slightly different from the density , which depends on the size of the domain and could be locally nonuniform , but it should correlate well with it .
we have not attempted to verify this , since it would require a separate study . to quantify the mixing in our two model flows we calculate numerically their topological entropies .
figure [ fig : lengths ] shows the interface length @xmath60 for the two flows in figures [ fig : singly ] and [ fig : doubly ] .
the curves quickly reach an exponential growth regime , though there is clearly some variation in growth rate during each period of the flow . to allow for this
we calculate the slope of the curves from a linear least - squares fit , ignoring transient data in the first period of the flow .
there is clearly some subjectivity about how such a fit is performed , so we claim our slopes are accurate to within ten percent .
we find that the @xmath58 flow has an entropy of @xmath77 , so the lower bound of @xmath46 derived in section [ sec : periodic ] , based only on the braiding of the stirrers , is around @xmath78 sharp . for the doubly - periodic braid @xmath54 our numerically computed entropy is around @xmath79 ,
meaning that the lower - bound of @xmath57 is at least @xmath78 sharp .
the sharpness of the bounds indicates that the braiding motion of the stirrers in our two model flows are responsible for almost all of the chaos , and only a small amount of chaos is due to secondary effects such as periodic orbits .
the fact that fluid particles themselves act as topological obstacles to flow in two dimensions @xcite means that rigorous topological entropy lower bounds can be computed even when stirring apparatus is not present to force a braiding motion . to provide an illustration we consider the well - known sine flow @xcite , which has not been examined from such a topological perspective .
the doubly - periodic sine flow is defined on the unit square @xmath63 .
we consider , for simplicity , the case where the velocity field has a temporal period of unity , with the velocity field given by @xmath80 for @xmath81 and @xmath82 for @xmath83 .
it is easy to verify that in this case the trajectories of the initial points @xmath84 , @xmath85 , @xmath86 , @xmath87 , @xmath88 , @xmath89 , @xmath90 and @xmath91 are all periodic , with period @xmath16 .
figure [ fig : sineexamples ] shows the result of advecting the initial condition described in section [ sec : numerics ] according to the sine flow .
it is easy to see from this numerical simulation that the trajectories of these points form a non - trivial braid .
one way of writing this braid is @xmath92 .
note how highly - curved folds of the interface form around some of the periodic points , showing that these points really do act like small stirrers !
this can occur because these are parabolic points with a singular unstable manifold , allowing material lines to fold around the point .
the nature of these points and the folding behaviour will be addressed in future work . in figure
[ fig : sinetrack ] we show the construction of the train - track for this braid , calculated in the same way as described in section [ sec : periodic ] .
the snapshots of material lines in figure [ fig : sineexamples ] are again invaluable in determining the train - track .
the edge to edge - path mapping is @xmath93 with the corresponding transition matrix @xmath94\ ] ] having a spectral radius of @xmath95 and topological entropy @xmath96 .
advection of material lines in this flow can be computed quickly using a map , rather than solving the full flow numerically , therefore we can compute @xmath60 for larger @xmath61 and subsequently estimate the entropy more accurately at around @xmath97 . so the train - track bound is a reasonable @xmath98 sharp , to within a percent .
this bound is slightly less sharp than for the examples shown in section [ sec : numerics ] .
perhaps this is to be expected since we did not force the flow to execute a particular ` good ' braid .
instead , we chose periodic orbits that were easily identified analytically .
however , the bound could easily be improved by including further periodic points in the train - track calculation @xcite .
we leave this for a future study .
we have shown how to compute lower bounds on the topological entropy of a two - dimensional time- and spatially - periodic fluid flow using the ` iterate and guess ' approach for constructing train - tracks .
for both one and two periodic directions this approach may be applied directly by considering graphs embedded on a cylinder or torus . an alternative approach may be used for singly - periodic flows , where a conformal transformation leads to a braid in an annular domain with equal topological entropy . in the latter approach , performing the conformal mapping introduces an additional flow obstacle which leads to a braiding motion with an additional , fictitious , stationary stirrer .
this confirms that in a spatially periodic flow it is possible to create topological chaos with just two actual stirrers , rather than the three required in a bounded or non - periodic infinite flow @xcite .
the possibility of topological chaos also follows from the existence of badly - ordered periodic orbits of the circle map associated with the stirrer motion , which implies positive entropy @xcite .
note that it does not follow that tc can be produced with a single stirrer on a doubly periodic domain .
the toroidal braid group with one stirrer is simply @xmath99 ( no @xmath100 s ) .
these two elements commute with each other @xcite , so it is not possible to generate a non - trivial braid .
intuitively , with only one stirrer there is nothing to wrap the stirrer around and hence it is impossible to tangle fluid lines around anything .
flows with one stirrer can of course still be chaotic , with a finite topological entropy , but topological considerations with just one stirrer trajectory reveal nothing .
the train - track approach actually provides more information than just the growth rate of material lines .
the entries in the transition matrix eigenvector corresponding to the largest eigenvalue reveal where edges in the train - track are accumulated at large times . in general the entries in this vector are different from each other and account for the spatial inhomogeneity in the relative number of striations between the stirrers , analogous to the situation described by alvarez _ et al . _
@xcite . in model stokes flow simulations we have shown for the braids @xmath101 and @xmath102 the computed entropy bounds are around @xmath78 sharp .
these bounds are surprisingly good given that they are based only on the braid performed by the imposed stirrer motions , however they could be made even sharper by including information about other , secondary , periodic structures in the flow @xcite .
we found that the topological entropy of braids on the cylinder and torus was quite high .
this is not surprising : the periodic boundary conditions offer many more opportunities for complex braiding motions than the plane or a bounded domain , thus causing material lines to grow rapidly .
using the methods presented here , it was relatively simple to find braids in the sine flow that have positive topological entropy .
the fact that this topological entropy was @xmath98 sharp when compared to the measured topological entropy of the sine flow means that much of the observed chaos is embodied in these orbits .
all these orbits are parabolic , which allows folding of material lines around them an important mechanism for chaos .
thanks to toby hall for help with train - tracks , to igor mezi for helpful discussions , and to an anonymous referee for clarifying several aspects of thurston nielsen theory .
this work was funded by the uk engineering and physical sciences research council grant gr / s72931/01 .
p. boyland , `` an analog of sharkovski s theorem for twist maps , '' in _ hamiltonian dynamical systems ( boulder , co , 1987 ) _ , vol .
81 of _ contemp .
providence , ri , 1988 ) , pp . | topologically chaotic fluid advection is examined in two - dimensional flows with either or both directions spatially periodic .
topological chaos is created by driving flow with moving stirrers whose trajectories are chosen to form various braids .
for spatially periodic flows , in addition to the usual stirrer - exchange braiding motions , there are additional topologically - nontrivial motions corresponding to stirrers traversing the periodic directions .
this leads to a study of the braid group on the cylinder and the torus .
methods for finding topological entropy lower bounds for such flows are examined .
these bounds are then compared to numerical stirring simulations of stokes flow to evaluate their sharpness .
the sine flow is also examined from a topological perspective . |
consider an ising ferromagnet consisting of an assembly of @xmath0 spins , each placed at a site of a @xmath1-dimensional euclidean lattice .
let lowercase italic letters , @xmath2 , @xmath3,@xmath4 , denote such sites , and @xmath5 , @xmath6 , @xmath4 , the relevant spin variables , so that the energy pertaining to a given configuration @xmath7 has the form @xmath8 where the symbol @xmath9 in the second sum restricts it to the sites @xmath3 adjacent to @xmath2 .
as it is well known , the ising model@xcite is amenable of an exact solution on simple two dimensional lattices @xcite , whereas for more complex lattices or higher dimensions the study of its thermodynamic properties strongly relies on numeric simulations .
according to the so - called _ dynamic monte carlo method _ , the thermodynamic canonical average of a generic observable @xmath10 is obtained as a simple algebraic average over the markov chain of configurations produced by a suitable stochastic algorithm @xcite .
indeed each configuration is obtained from the previous one so that , in the asymptotic limit , the probability @xmath11 that a given configuration @xmath12 occurs at temperature @xmath13 is proportional to the boltzmann canonical factor , @xmath14 $ ] , independent of the initial configuration .
one of the most commonly exploited algorithms is the glauber s single - spin - flip algorithm @xcite . according to glauber s prescription two subsequent configurations
can differ at most by the value of a single spin variable .
more precisely the spin of the configuration @xmath12 relevant to the site @xmath15 can flip with probability @xmath16}\ ] ] where @xmath17 is the energy variation consequent to the process .
glauber s dynamics is completely defined by eq .
( [ e : gp ] ) together with a prescription for updating the spin system . the latter is commonly chosen in view of a computational optimization and typically consists in an unphysical sweep along parallel lattice lines@xcite . as we mentioned above such a dynamics has a computational rather than physical origin , having been devised as an efficient way to give estimates of the canonical averages .
nevertheless it is usually given a physical interpretation @xcite : the energy variation consequent to the spin - flips realizing the evolution of the system occur due to the coupling of the spin degrees of freedom to a heat - bath at temperature @xmath13 .
in the last decade a number of new dynamics for the ising model were introduced@xcite , where the energy variations are typically required to occur uniformly throughout the whole sample . here
we introduce an alternative dynamics exhibiting the diffusive behaviour of a random walk in an evolving landscape generated by the energy of the spin couplings in the real space .
the spin - flips are induced by a random walker hopping across the sites of the underlying discrete structure .
the motion of such a random walker is affected by the spin interaction , being biased towards those sites where a spin - flip is energetically more favorable .
it should be remarked that a relaxation dynamics occurring under the action of random walkers was introduced by creutz@xcite , and subsequently exploited by many authors .
however our walkers act quite differently from creutz s _ demons_. indeed the latter diffuse freely , unbiased by the magnetic configuration of the system , and , in order to simulate the micro - canonical ensemble , they induce a spin flip only if they can afford it , according to the individual energy they are endowed with . in fact our model in inspired by the physical behaviour of manganites @xcite , such as @xmath18 , where the spin dynamics is determined by the presence of diffusing excitations interacting with the magnetic degrees of freedom . in particular , for the manganites
the excitations are given by the charged carriers , electrons and vacancies , present in the non - stoichiometrical compounds .
obviously , this model is not intended to be a phenomenological description of such complex systems , but it aims to evidence the influence of diffusing excitation interacting with spins on the thermodynamic behaviour of the paradigmatic model of all magnetic phase transitions .
the analysis of the diffusive dynamics is carried out from the numerical point of view .
all the simulations illustrated in the present paper refer to two - dimensional arrays of spins , so that the observed features of the statistical model can be compared to the analytically known results pertaining to the canonical ising model .
the plan of the paper is as follows : in section [ s : dd ] we describe and briefly discuss the diffusive algorithm .
section [ s : nces ] and [ s : cb ] are devoted to the analysis of the limit situation where the dynamics is realized by a single walker . in the former we provide numerical evidences that the magnetic system is driven to a thermodynamically well - behaved steady state which differs from the canonical equilibrium state in a non trivial - manner .
since a magnetic phase transition is still observed , the consequent critical behaviour is analyzed in section [ s : cb ] .
the estimated critical exponents do not deviate significantly from the values pertaining to the canonical ising model . in section [ s : de ]
the results of simulations where the system evolves subject to more than a single walker are analyzed .
section [ s : cp ] consists in our conclusions .
as we mentioned in the previous section the relaxation dynamics we propose is realized by means of random walkers diffusing through the sites of the ising system . the probability that a walker located at site @xmath2 hops on site @xmath3 and flips the relevant spin , where @xmath3 is one of the @xmath19 neighbours of @xmath2 ,
is given by @xmath20 hence , for any configuration @xmath21 of the ising system and for any value of the temperature @xmath13 , the diffusion of the walker is biased towards those sites where a spin - flip is more likely to occur according to glauber s probability , eq .
( [ e : gp ] ) .
note that equation ( [ e : pij ] ) implicitly yields a probability @xmath22 that the walker does not move .
if this is the case the magnetic configuration of the system @xmath21 remains unchanged as well . according to equations ( [ e : pij ] ) and ( [ e : pii ] ) the evolution of the probability @xmath23 that , at time step @xmath24 , the magnetic system is in the configuration @xmath21 while the walker is located at site @xmath2 , is governed by the master equation @xmath25,\end{aligned}\ ] ] where @xmath26 is a shorthand notation for the configuration obtained flipping the spin @xmath27 of @xmath21 .
unfortunately the master equation approach is of very little use in analyzing the long - time behaviour of this process .
actually in the canonical heat - bath case the master equation was exploited to build a dynamics driving the system to a known asymptotic probability .
this was obtained by imposing the quite restrictive _ detailed balance condition_. conversely , we introduced an evolution algorithm and our aim is the study of the resulting asymptotic state of the ising system , if any .
hence we carry out our analysis mainly by means of numerical simulations . for a better comparison we focus on the two - dimensional system , where most of the results pertaining to the canonical ising model are exactly known .
hence all of our simulations refer to an ising system consisting of @xmath28 spins placed at the sites of a two - dimensional square lattice . in order to avoid time consuming procedures dealing with the walkers bumping into the borders of the system we adopt periodic boundary conditions . in the next two sections we focus on the situation where the evolution of the ising system is realized by a single walker .
this is a limit case , in that , for a reasonably macroscopic system , it corresponds to a vanishing density of walkers .
the results pertaining to larger densities are discussed in section [ s : de ] .
in the following we give numerical evidence that the spin - flip dynamics realized by a single walker drives the magnetic system to a thermodynamically well - behaved steady state which differs from the canonical equilibrium state in a non trivial manner . at @xmath29 .
all the measurements were carried out in the stationary regime .
the error - bars represent the fluctuations about the average values .
, width=325 ] at @xmath29 .
all the measurements were carried out in the stationary regime .
the error - bars represent the fluctuations about the average values .
note that despite the symmetry of the hamiltonian the measured value for the specific magnetization is non - zero .
similar to the canonical case the ising system displays a phase transition.,width=325 ] all the simulations we performed show that , at a given value of the external parameter @xmath13 the system eventually reaches a steady state , characterized by a well defined value of the ( time ) average of the macroscopic observables .
these features are clearly recognizable in figures [ f : nrgs ] and [ f : mags ] , where the average values of the specific energy @xmath30 and magnetization @xmath31 are plotted for systems with different sizes , at a fixed value of the temperature parameter . furthermore ,
as one would require , the entire sample and any reasonably macroscopic portion of it are characterized by the same specific values of the thermodynamic observables .
as it is clearly shown in figures [ f : nrgf ] and [ f : magf ] , the fluctuations about the average value of the specific thermodynamic observables exhibit the expected scaling behaviour , decreasing as the inverse square root of the system size . at @xmath29 .
the slope of the linear fit ( dotted line ) of the measured data ( @xmath32 ) , @xmath33 , is in good agreement with the expected value , @xmath34 .
all the measurements were carried out in the stationary regime.,width=325 ] at @xmath29 .
the slope of the linear fit ( dotted line ) of the measured data ( @xmath32 ) , @xmath33 , is in good agreement with the expected value , @xmath34 .
all the measurements were carried out in the stationary regime.,width=325 ] figure [ f : spmag ] shows the average values of the specific energy and magnetization at different values of the temperature parameter , @xmath35 .
note that , similar to the canonical ising model , the system exhibits a magnetic phase transition .
400 ising array . filled circles ( @xmath32 ) : specific magnetization ; open circles ( @xmath36 ) : specific energy.,width=325 ]
we verified that the long time behaviour of the macroscopic observables of the system is uniquely determined by the only external parameter characterizing the dynamics , namely the temperature @xmath13 .
in particular different choices for the initial configuration of the magnetic system do not affect the asymptotic behaviour of the system , except for a possible trivial bias on the orientation of the spontaneous magnetization .
similar to what happens in the heat - bath case , both of the possible orientation of the specific magnetization are in principle equally likely , but a strongly magnetized initial configuration is very likely to evolve in a stationary state exhibiting a spontaneous magnetization along the same direction . when the initial configuration has no net magnetization there is no bias on the direction of the spontaneous magnetization exhibited by the system below the critical temperature . in this situations the formation of macroscopic domains exhibiting opposite net magnetization is observed ( see fig.[f : domains ] ) . in the long time regime one of the domains eventually prevails against the others . [ cols="^,^ " , ] as we already mentioned the system exhibits a magnetic phase transition . from fig .
[ f : spmag ] and [ f : domains ] , it is clear that the critical temperature for the diffusive dynamics is significantly larger than the value pertaining to the canonical ising model . such value , @xmath37 is analytically known and the heat - bath dynamics yield quite satisfactory numerical estimates of it . as we will see in the next section ,
accurate estimates yield the value @xmath38 for the critical temperature of an ising system evolving under the action of a single walker .
however , such a quantitative difference can not be accounted for by means of a simple rescaling of the temperature .
indeed if the diffusive dynamics acted as heat - bath dynamics with a rescaled temperature , @xmath39 , the generic configuration @xmath12 would occur with a probability @xmath40 $ ] in the asymptotic regime @xcite .
hence the joint probability for the occurrence of a configuration such that @xmath41 and @xmath42 would be of the form @xmath43 where @xmath44 and @xmath45 last equality of eq .
( [ e : partf ] ) was obtained making use of eq .
( [ e : combf ] ) .
now note that , according to eqs .
( [ e : jointp ] ) and ( [ e : combf ] ) , @xmath46 hence , if the rescaling hypothesis was true , the plot of @xmath47 versus @xmath48 at a fixed value of the specific energy would not depend on the temperature : the curves pertaining to the same specific energy at different values of the temperature @xmath13 would overlap .
this can be verified with great precision in the case of the heat - bath dynamics @xcite the numeric estimates of these curves for two different values of the specific energy @xmath49 are plotted in figures [ f : dos1 ] and [ f : dos2 ] . for a @xmath50 ising array subject to the diffusive dynamics .
all of the curves refer to the specific energy @xmath51 .
filled circles ( @xmath32 ) : @xmath52 ; open circles ( @xmath36 ) : @xmath53 ; crosses ( @xmath54 ) : @xmath55 ; dotted line : heat - bath dynamics ( temperature independent).,width=325 ] for a @xmath50 ising array subject to the diffusive dynamics .
all of the curves refer to the specific energy @xmath56 .
filled circles ( @xmath32 ) : @xmath53 ; open circles ( @xmath36 ) : @xmath55 ; crosses ( @xmath54 ) : @xmath57 ; dotted line : heat - bath dynamics ( temperature independent).,width=325 ] the fact that the curves pertaining to the diffusive dynamics at different temperatures are well distinct from one another and from the ( temperature independent ) curve characterizing the canonical ising model proves that the differences between the diffusive and the heat - bath dynamics can not be accounted for by means of a simple rescaling of the temperature . in summary the steady state
asymptotically reached by an ising system subject to the action of the diffusive dynamics is thermodynamically well - behaved , and yet it is non - trivially different from the canonical equilibrium of the ising model .
hence we will refer to the steady state of the diffusive dynamics as a _ non - canonical equilibrium state_. in the following section we analyze the critical behaviour of the ising system subject to the diffusive dynamics .
as we already observed in the previous section , an ising system subject to a single diffusive walker exhibits a magnetic phase transition at a finite value of the temperature ( see figure [ f : spmag ] ) .
similar to what happens with the ising model , this phenomenon is accompanied by a singular behaviour of the thermodynamic functions .
, of the magnetization for an ising system subject to the diffusive dynamics .
the measures were performed on a system consisting of @xmath58 spins .
the error bars denote the standard deviation of the measures .
the dotted line is the best fit , @xmath59 .
the estimated values for the critical temperature and for the exponent are @xmath60 and @xmath61 , respectively .
the latter is consistent with the relevant critical exponent of the canonical ising model , @xmath62 .
the vertical dashed line indicates the critical temperature.,width=325 ] in figure [ f : beta ] accurate data for the magnetization of an ising system subject to the diffusive dynamics are plotted .
the error bars represent the standard deviations about the average values . similar to the canonical case ,
the data show a good agreement with a critical behaviour of the form @xmath63 and @xmath64 , respectively .
the latter result is in good agreement with the relevant critical exponent of the two - dimensional canonical ising model , @xmath65 . as we already discussed in the previous section ,
the critical temperature is appreciably larger than the canonical value , @xmath66 .
figure [ f : sph ] shows the specific heat @xmath67 for a @xmath68 ising array subject to the diffusive dynamics as a function of the temperature @xmath13 .
the vertical dashed lines are placed at the critical value of the temperature .
the dotted curves refer to functions of the form @xmath69 , and they fit quite satisfactorily the data . hence , similar to the canonical case
, there is a strong signature of a logarithmic divergence of the specific heat at the critical temperature . ) .
the dotted curves fitting the data are of the form @xmath69 .
the vertical dashed line indicates the estimated value of the critical temperature , @xmath70 .
the diamonds ( @xmath71 ) refer to the quantity @xmath72 defined in eq .
( [ e : flc ] ) . , width=325 ] figure [ f : gamma ] shows a log - log scale plot of the quantity @xmath73\ ] ] versus @xmath74 .
the data are consistent with a power law of the form @xmath75 , and the slope @xmath76 of the linear fit is in good agreement with the critical exponent governing the behaviour of the same quantity in the of the canonical case , @xmath77 .
we recall that for a system at canonical equilibrium the fluctuation - dissipation theorem yields the relations @xmath78 and @xmath79 , where @xmath80.\ ] ] and @xmath81 is the magnetic susceptibility of the system . in our case
this is not necessarily true , since the equilibrium distribution for a diffusive dynamics at the temperature @xmath13 is not proportional to the canonical boltzmann factor @xmath82 $ ] , as we discussed in the previous section .
hence the critical behaviour of two quantities canonically related by a fluctuation - dissipation relation should be explicitly analyzed and compared .
a preliminary step in this sense is shown in figure [ f : sph ] , where the diamonds represent the estimates of the quantity @xmath72 defined in eq .
( [ e : flc ] ) .
these results seem to indicate that the diffusive character of the dynamics does not produce large deviations from the fluctuation - dissipation relation .
of course this result needs to be checked through the analysis of the response of the system to an external magnetic field .
however we mention that other studies @xcite of ising systems subject to dynamics which do not yield the canonical equilibrium state actually estimate the critical exponent @xmath83 in terms of the quantity @xmath84 defined in eq .
( [ e : chif ] ) .
( defined in terms of the magnetic fluctuations , see eq .
( [ e : chif ] ) ) versus @xmath74
. the estimated slope of the straight line fitting the data is @xmath85 .
this value is consistent with the critical behaviour of the magnetic susceptibility for the two - dimensional canonical ising model , characterized by the critical exponent @xmath77.,width=325 ] for temperatures sufficiently far from the critical value the correlation function @xmath86 exhibits an exponential decrease as the distance between the relevant sites , @xmath87 , increases : @xmath88\ ] ] as it is shown in fig . [
f : xit ] the correlation distance @xmath89 features a maximum very close to the critical temperature .
the fact that this maximum suggests a critical temperature slightly larger than the value previously ( and more accurately ) estimated must be ascribed to finite size effects .
we further mention that the correlation lengths are roughly consistent with a power - law divergence of the form @xmath90 . ising array subject to the diffusive dynamics . due to finite size effects
, the critical behaviour is exhibited for temperature slightly higher than the estimated critical value , @xmath91.,width=325 ] as it is well known @xcite , for temperatures very close to the critical value , and hence for correlation lengths very large compared to the lattice constant , the correlation function of the canonical ising model is characterized by a power - law behaviour of the form @xmath92 .
the relevant critical exponent is defined to be @xmath93 , where @xmath1 is the dimensionality of the ising array .
for the two - dimensional canonical ising model its value is exactly known to be @xmath94 .
such behaviour is expected to be observable in the distance range @xmath95 . versus the inter - spin distance , @xmath96 .
the data appreciably deviate from a power law behaviour .
the dotted line refer to a function of the form defined by eq .
( [ e : cffit ] ) .
the exponent @xmath97 is quite close to the critical exponent @xmath98 .
more precisely we estimate @xmath99 .
the same fit satisfactorily applies to the data relevant to @xmath100 , where the exponent @xmath101 is once again in good agreement with the canonical value.,width=325 ] as it is shown in figure [ f : eta ] the data yielding the largest correlation length are not consistent with a pure power law behaviour , not even for relatively small distances .
this suggests that the relevant temperatures , @xmath102 is not sufficiently close to the critical value .
moreover it should be recalled that the finite size of the system can introduce appreciable deviations from the theoretical prediction , which is strictly true in the thermodynamic limit .
nevertheless we mention that a function of the form @xmath103 fits the data quite satisfactorily over a wide range of distances .
the estimate for the exponent , @xmath104 is remarkably close to the critical exponent @xmath105 of the two - dimensional ising model .
a fit of the same kind satisfactorily applies to the data relevant to the temperature @xmath100 , and it gives a critical exponent @xmath106 once again in good agreement with the canonical value .
these results could be an indication of the fact that , in the suitable regime , the correlation function @xmath107 actually exhibits a power law behaviour characterized by the same critical exponent as the two - dimensional canonical ising model .
the results illustrated so far suggest that a two - dimensional ising ferromagnet subject to the diffusive dynamics exhibits a critical behaviour belonging to the same universality class as the canonical case . before concluding the present section we mention a further result which has no counterpart in the canonical case . indeed , in the diffusive case , it is possible to relate the configuration of the magnetic system to the position of the random walker causing its evolution .
more precisely it is possible to measure to what extent the presence of the random walker at a given site @xmath2 and the local magnetization at a given site @xmath3 influence each other .
numerical estimates of such correlation , which we illustrate elsewhere @xcite , show that it exhibits a critical behaviour in correspondence of the critical temperature .
the results illustrated in the previous two sections refer to the case of an ising system subject to the action of a single walker . here
we analyze the results of simulations where the evolution of a square ising array consisting of @xmath0 spins is realized by @xmath108 non interacting walkers .
the simulation is initialized giving each walker a randomly chosen position .
two or more walkers are allowed to occupy the same site of the system .
actually this is very likely to happen for a sufficiently large density of walkers @xmath109 .
an elementary step of the simulation consists in the application of the algorithm described in section [ s : dd ] to all of the walkers , according to a fixed progressive sequence .
when its turn comes , a given walker acts as if the other ones were not present .
a qualitative argument leads us to expect that in the case of an infinite density of walkers the results pertaining to the canonical ising model are recovered .
consider a very large density of walkers , such that @xmath110 .
in this situation the effect of a single step of the simulation is very similar to the action of the glauber s dynamics with a random sequence update .
indeed the probability that a walker flips a given spin @xmath5 is given by the product of the probability @xmath111 that it is located at one of the four neighbours of site @xmath2 times the hopping probability , @xmath112 .
but @xmath113 is exactly the probability that the spin @xmath5 is flipped at any step of the glauber s heat - bath dynamics with random site update .
hence , in the limit case of an infinite density of walkers , the diffusive dynamics should necessarily drive the system to the same asymptotic state as the heat - bath dynamics , thus reproducing the results of the canonical ising model .
the same should be clearly true for a sufficiently large density of walkers , provided that the consequent number of elementary steps of the heat - bath dynamics drive the system sufficiently close to the canonical equilibrium state .
ising array .
the horizontal dotted lines are placed at the critical temperatures of the diffusive dynamics with a single walker ( @xmath114 ) and at the critical temperature of the canonical ising model ( @xmath115).,width=325 ] the numerical simulations we performed evidenced that the diffusive dynamics drives the ising system to a thermodynamically well behaved stationary state , so that the average values of the specific energy and magnetization are characterized only by the temperature , @xmath13 , and the density of walkers , @xmath116 . in particular they confirmed the qualitative argument discussed above : the results pertaining to the canonical ising model at a given temperature @xmath13 were recovered for a sufficiently large density of walkers .
more generally we verified that the average value of a macroscopic observable relevant to a finite density lies within the interval between the values pertaining to the limit situations , namely the vanishing density limit @xmath117 analyzed in sections [ s : nces ] and [ s : cb ] , and the canonical ising model .
the influence of the density of walkers is synthesized in figure [ f : ntc ] , where the estimated values of the critical temperature are plotted .
note that the larger variation occurs for densities between @xmath118 and @xmath119 .
in this paper we exploited the basic microscopic model of diffusion , namely the random walk , to build a dynamics for the ising ferromagnet which retains the diffusive character of the thermal motion of a localized excitation . in this framework
the density of walkers is related both to the intensity and spatial extent of the coupling of the spins to the thermal degrees of freedom of the system , and hence to the velocity of the thermal dynamics .
the larger is the density of walkers , the faster is the dynamics and the closer to the uniform heath - bath are its effects .
even if our model is rather simplified with respect to the real physical systems which inspired it , it exhibits new important qualitative features which should be relevant even in more complex and realistic situations
. therefore it would be of great interest to verify whether the diffusive thermal dynamics gives a satisfactory description for the slow evolution of a magnetic system with hopping excitations .
this would of course require the extension of the study of its effect on the more realistic three dimensional lattice . in view of recognizing the action of a diffusive dynamics in a real system
, the results of section [ s : cb ] suggest that the study of the critical behaviour is not particularly significant .
indeed the latter does not allow to discriminate among the canonical and the diffusive dynamics .
on the other hand the results illustrated in section [ s : de ] indicate that the velocity of the non - equilibrium dynamics ( related to the degree of coupling of the spins to the thermal vibrations , i.e. to the density of excitations induced in the system by an external source ) significantly affects the critical temperature of the system .
hence the same system subject to diffusive thermal dynamics with different velocity should display different critical temperatures .
furthermore we remark that fig .
[ f : domains ] suggests that the velocity of the dynamics influences the shape and growth of the macroscopic magnetic domains .
in particular a very slow diffusive dynamics gives rise to domains with smoother boundaries .
this feature suggests that the final shape of the magnetic domains within a physical system possibly governed by a diffusive thermal dynamics could be controlled by a fine tuning of the density of excitations mediating the coupling to the thermal degrees of freedom
. a detailed study of domain growth in the diffusive dynamics will be the subject of a forthcoming paper @xcite .
r. j. glauber , j. math .
phys . * 4 * , 294 ( 1963 ) .
this prescription is often referred to as _ typewriter _ update sequence . on a bipartite lattice
it may prove numerically efficient to update each sublattice according to a typewriter sequence .
the resulting sequence is referred to as _ checkerboard _
update sequence .
see e.g. t. vojta and m. schreiber , phys .
e , * 58 * , 7998 ( 1998 ) . see e.g a.m. mariz , f. nobre and c. tsallis , phys .
b * 49 * 3576 ( 1994 ) and references therein .
wang , eur .
j. b * 8 * , 287 ( 1999 ) .
heringa and w.j .
blte , phys .
e * 57 * 4976 ( 1998 ) .
m. creutz , phys .
lett . * 50 * , 1411 ( 1983 ) ; g. bhanot , m. creutz and h. neuberger , nucl .
phys b * 235 * , 417 ( 1984 ) e. dagotto , t. hotta and a. moreo , phys .
rep . * 344 * , 1 ( 2001 ) . in that case
the rescaling is trivial : @xmath120 .
h. w. blte , j. r. heringa , a. hoogland , and r. k. p. zia , j. phys .
a : math . gen . * 23 * , 3799 ( 1990 ) .
l. p. kadanoff , il nuovo cimento b , xliv , 5152 ( 1966 ) .
p. buonsante , r. burioni , d. cassi and a. vezzani , _ in preparation_. | we introduce a thermal dynamics for the ising ferromagnet where the energy variations occurring within the system exhibit a diffusive character typical of thermalizing agents such as e.g. localized excitations .
time evolution is provided by a walker hopping across the sites of the underlying lattice according to local probabilities depending on the usual boltzmann weight at a given temperature . despite the canonical hopping probabilities the walker drives the system to a stationary state which is not reducible to the canonical equilibrium state in a trivial way .
the system still exhibits a magnetic phase transition occurring at a finite value of the temperature larger than the canonical one .
the dependence of the model on the density of walkers realizing the dynamics is also discussed .
interestingly the differences between the stationary state and the boltzmann equilibrium state decrease with increasing number of walkers .
pacs numbers : 2.70.uu , 2.70.tt , 5.50.+q , 5.10.ln |
cataclysmic variables ( cvs ) are short period binary systems in which a ( typically ) late main - sequence star ( the secondary ) transfers material via roche - lobe overflow to a white dwarf primary star . for an excellent review of cvs ,
see @xcite .
although cvs are largely observed to study the fundamental astrophysical process of accretion , it is the roche - lobe filling secondary stars themselves that are key to our understanding of the origin , evolution and behaviour of this class of interacting binary . in particular , the magnetic field of the secondary star
is thought to play a crucial role in the evolution of cvs driving cvs to shorter orbital periods through _ magnetic braking _ ( e.g. @xcite , @xcite , @xcite , @xcite ) .
furthermore , the transition of the secondary star to a fully convective state and the supposed shutdown of magnetic activity that this transition brings is also invoked to explain the period gap the dearth of cvs with orbital periods between @xmath11 23 hours . on more immediate timescales , magnetic activity on the secondary stars are also thought to explain variations in cv orbital periods , mean brightnesses , mean outburst durations and outburst shapes ( e.g. @xcite ; @xcite ; @xcite ) .
that the secondary star and its magnetic field should have such a large and wide - ranging impact on cv properties should come as no surprise the secondary star essentially acts as the fuel reserve that powers these binaries . therefore , an understanding of the magnetic field properties of the secondary stars in cvs ( e.g. spot sizes , distributions and their variation with time ) is crucial if we are to understand the behaviour of these binaries .
in addition , detailed studies of the rapidly rotating secondary stars in cvs can also provide tests of stellar dynamo theories under extreme conditions .
for example , questions regarding the impact of tidal forces on magnetic flux tube emergence ( e.g. @xcite ) and its effect on differential rotation ( e.g. @xcite ) are particularly pertinent ( see @xcite for a discussion ) . despite this ,
until recently there had been little direct observational evidence for magnetic activity in cvs .
@xcite used tio bands to infer the presence of spots on the secondary star in ss cyg and estimated a spot filling factor of 22 per cent .
unfortunately , this technique does not allow the surfaces of these stars to be imaged and hence the spot distributions could not be ascertained .
most recently , @xcite used roche tomography to map the starspot distribution on a cv secondary ( ae aqr ) for the first time . in @xcite
we estimated that starspots covered approximately 18 per cent of the northern hemisphere of ae aqr .
the roche tomogram of ae aqr also showed that starspots were found at almost all latitudes , although there was a relative paucity of starspots at a latitude of @xmath1140@xmath0 .
furthermore , we found that , in common with doppler images of single rapidly rotating stars , ae aqr also displayed a large high - latitude spot . in this work
we continue our series of papers on roche tomography ( see @xcite , @xcite , @xcite ) by mapping starspots on the long period ( 0.61-d ) dwarf nova bv cen for the first time .
we also report on the serendipitous discovery of the binarity of hd220492 .
simultaneous spectroscopic and photometric observations were carried out over 3 nights on 2004 july 810 .
the spectroscopic data were acquired using the 6.5-m magellan clay telescope and the simultaneous photometry was carried out using the carnegie institution s henrietta swope 1.0-m telescope .
both telescopes are situated at the las campanas observatory in chile .
the spectroscopic observations of bv cen were carried out using the dual - beam magellan inamori kyocera echelle spectrograph ( mike see @xcite ) . the mit lincoln labs ccd-20 chip with 2046 @xmath12 4096 pixels was used in the blue channel , and the site st-002a chip , again with 2046 @xmath12 4096 pixels , was used in the red channel .
the standard setup was used , allowing a wavelength coverage of 3330
5070 in the blue arm and 4460 7270 in the red arm , with significant wavelength overlap between adjacent orders . with a slit width of 0.7 arcsec , a spectral resolution of around 38,100 ( @xmath117.8 km s@xmath10 ) and 31,500 ( @xmath119.5 km s@xmath10 )
was obtained in the blue and red channels , respectively .
the chips were binned 2@xmath122 resulting in a resolution element of @xmath11 2.3 binned pixels in the red arm and 2.6 binned pixels in the blue arm for our chosen slit .
the spectra were taken using 400-s exposure times in order to minimise velocity smearing of the data due to the orbital motion of the secondary star .
comparison thar arc lamp exposures were taken every @xmath1150 minutes for the purposes of wavelength calibration . with this setup
we obtained 63 usable spectra in each arm .
since the main goal of the magellan run was to observe ae aqr , we were restricted to 3-hour windows each night to observe bv cen . over our allocated 4 nights
this would have allowed over 80 per cent of the orbit of bv cen to be observed , but unfortunately we lost the final night due to bad weather .
other than that , the seeing was typically 0.60.7 arcsec on the first night and 0.9 arcsec on the next two nights , with occasional degradation to 1.5 arcsec .
the peak signal - to - noise of the blue spectra ranged from 29 56 ( typically @xmath1150 ) in the blue arm , and from 38 76 ( typically @xmath1165 ) in the red arm .
table [ table : log ] gives a journal of the observations .
it should be noted that when using mike it is not possible to change the slit orientation on the sky . in order to compensate for this , and reduce atmospheric dispersion across the slit , mike is mounted at a 30@xmath0 angle to the naysmith platform .
this means that at zenith distances greater than @xmath1150@xmath0 dispersion across the slit becomes significant . to avoid this , all observations of bv cen were carried out at low airmasses and no exposure of bv cen was carried out for an airmass above 1.39 ( zenith distance @xmath13 44@xmath0 ) . [ cols="<,<,^,^,^,^,<",options="header " , ]
in addition to the many starspot signatures that are visible in bv cen s trailed spectra ( see fig [ fig : trails ] ) , a curious narrow feature is also evident between phases 0.328 and 0.366 on the blue edge of the profile .
this feature appears as a narrow continuation of the main track through the profiles during this block of observations , but lies outside the stellar absorption profile and therefore can not lie on the stellar surface .
closer inspection of the individual profiles ( see fig . [
fig : sling ] ) reveals a narrow , weak emission feature in the continuum of 5 lsd profiles . in order to confirm the reality of this feature , we have visually inspected the profiles
that were deconvolved from the blue and red spectra separately to see whether they are present in both sets of profiles . indeed
this feature is present ( albeit very weakly ) in both sets of profiles which makes it unlikely to be due to noise or a systematic effect arising during the lsd process .
this feature is also not due to contamination from lunar or solar light during the observations as this would appear as an absorption feature ( e.g. @xcite ) .
furthermore , moon - rise did not occur until at least 2 hours ( extending to 4 hours on the final night ) after observations of bv cen were concluded .
the fact that the feature is at zero velocity with respect to bv cen s systemic velocity of -22 km s@xmath10 rules out a ` terrestrial ' origin , which would be centred at 0 km s@xmath10 .
we are therefore confident that this feature is real .
since the emission feature lies outside the stellar line profile , and hence lies off the stellar limb , it can only be attributed to circumstellar material .
solar prominences appear as bright emission loops when they are viewed off the solar limb , and it is probable that we are also seeing a prominence structure on bv cen .
indeed , large prominences have been reported on other rapidly rotating stars such as ab dor ( e.g. @xcite ) and speedy mic ( @xcite ) and are often observed as transient absorption features passing through the doppler broadened h@xmath14 stellar line .
recently , however , analysis of vlt data of speedy mic by @xcite has revealed rotationally modulated emission outside of the stellar h@xmath14 line due to loops of emission seen off of the stellar disc , but which can also be associated with prominences seen to transit the stellar disc at other times . in addition , peculiar low - velocity emission features seen in ss cyg and ip peg during outburst have also been interpreted as ` slingshot prominences ' ( @xcite ) .
@xcite discovered the presence of highly - ionized low velocity - dispersion material located between the @xmath1 point and the centre of mass in am her which they attributed to a slingshot prominence .
similarly , triple - peaked h@xmath14 lines following the motion of the donor star in am her has been reported by @xcite and @xcite .
the authors interpreted these as long - lived prominences on the donor star and noted that one component was consistent with stellar activity lying vertically above the @xmath1 point .
the emission feature we see in bv cen appears stationary at @xmath110 km s@xmath10 within the binary frame .
this is in keeping with the stationary slingshot prominences seen in ss cyg and ip peg by @xcite , and the low - velocity emission observed in other cvs ( @xcite ) .
generally , we would expect the emission from prominences observed off the stellar limb to be weak and undetectable for cvs given their faintness . however , it is possible that the prominences are illuminated by the accretion light , causing prominences forming between the donor star and the white dwarf to become visible when otherwise they would be undetectable .
certainly , the low velocity of the emission suggests a position close to the centre - of - mass of the binary at a point between the donor star and white dwarf where such illumination is most likely .
given that prominences are normally only seen in lines that form above the photosphere ( e.g. the hydrogen balmer lines ) , observing them in the lsd profiles which are obtained from photospheric lines is unexpected under normal conditions .
we believe that the emission seen in bv cen s photospheric lines is due to excitation of these species within low density gas in the prominence due to the impact of irradiation .
thus , we suspect that not only does irradiation cause the prominence to become highly visible , but also causes it to be observable in some photospheric lines in which prominences in a normal , unirradiated environment would not normally emit . we have considered that the emission could be due to a wind launched from the accretion regions
. such a wind , however , would exhibit a radial velocity modulation due to the orbital motion of the primary star which we do not observe .
furthermore , presumably the velocities of material in such a wind would produce a far broader emission line than observed in bv cen .
for these reasons , and the fact that it kinematically matches previous observations of slingshot prominences in cvs , we prefer the interpretation that this feature is the result of an irradiated prominence .
we find that the emission feature in bv cen is very narrow , with a velocity width ( @xmath15 ) of @xmath11 10 km s@xmath10 .
using this width we can place an upper limit on the emission source size , @xmath16 .
following @xcite , we assume the prominence is co - rotating with the secondary star , so we can write , @xmath17 where @xmath18 and @xmath19 are the radial velocity amplitudes of the white dwarf and donor star , respectively , and @xmath20 is the orbital separation . for the parameters derived for bv cen in section [ sec : pars ] , this places an upper limit of 75,000 km .
naturally , the velocity dispersion is increased by instrumental resolution , thermal broadening , saturation broadening and turbulence within the prominence itself . for a 10,000 k prominence
the thermal doppler velocity of hydrogen is 12.9 km s@xmath10 ( e.g. @xcite ) . since the lines included in our lsd are mainly heavier elements such as fe and ca , the thermal broadening will be far smaller at @xmath112 km s@xmath10 .
solar prominences exhibit turbulent motions of several km s@xmath10 , and @xcite estimate a turbulent velocity of 5 km s@xmath10 for the prominences observed on speedy mic .
however , the dominant broadening mechanism is the @xmath119.5 km s@xmath10 instrumental resolution of our observations . considering these limitations , our estimated maximum source size of 75,000 km
should be viewed with caution as it is possible that the source is actually unresolved .
we have further analysed the lsd profiles to check whether or not we can see the zero - velocity prominence as it tracks across the stellar disc .
we were unable to positively identify any feature with zero - velocity between phases 0.3740.522 during the first night s data .
this may be due to the large emission bump feature that traverses the profile due to irradiation and/or starspots ( see section [ sec : maps ] ) which probably masks the prominence itself .
analysis of the second night s data , however , does reveal a narrow emission bump at zero - velocity between orbital phases 1.9742.058 .
this is quite conspicuous ( see fig . [
fig : sling ] ) as the emission feature moves across the profile in the opposite direction to the starspot features , and roche tomography is clearly unable to fit it .
this agrees with the picture of a prominence holding material near the centre - of - mass of the binary , and that at orbital phase @xmath110 we are effectively looking over the top ( or possibly under the bottom ) of the star at the prominence on the other side .
this , therefore , means that the prominence structure must be raised above or below the orbital plane in order for it not to be eclipsed by the donor star at this phase .
given the lack of eclipse and the parameters found for bv cen in section [ sec : pars ] , combined with the assumption that the prominence is located above the centre - of - mass of the binary ( the point of zero - velocity ) , we find that the prominence must lie at least 160,000 km above the orbital plane .
if we assume that the prominence lies below the centre - of - mass ( i.e. we are looking ` underneath ' the star around phase 0 ) then it lies out of the orbital plane by at least 2,400,000 km .
we feel that this latter case is highly unlikely .
first , a prominence below the orbital plane is far more likely to be eclipsed by the accretion regions around orbital phase @xmath110.35 , when it is clearly seen in the data .
second , if we assume that the prominence is only visible due to illumination from the accretion regions then a prominence below the orbital plane is located too far away from the irradiating source .
three nights of consecutive ( albeit interrupted ) observations allows a limited discussion of the evolution of the slingshot prominence observed on bv cen .
the prominence is certainly seen at the start of the first night ( @xmath21 = 0.3280.366 ) before it transits the stellar disc , where
it then becomes invisible .
we have assumed that , rather than the prominence disappearing at this point , its signature is lost in the complex structure present in the stellar line profile .
we are able to pick up an emission feature with the same position in velocity space again at the start of the second night ( @xmath21 = 1.9742.038 , see fig .
[ fig : sling ] ) . curiously
, this prominence feature then disappears after orbital phase 2.038 . since , unlike the first night , the prominence on the second night appears quite clearly in the middle of the stellar line profile it is , on this occasion , difficult to explain how the feature could suddenly be lost within the stellar line
this suggests that the disappearance may be due to rapid evolution of the prominence itself .
indeed , it does seem as though the emission feature weakens during the second night before its apparent disappearance ( see fig . [
fig : sling ] ) .
though the interpretation of a prominence rapidly evolving on timescales of hours is speculative , @xcite also find evidence of individual prominences evolving on timescales of @xmath119 hours on speedy mic .
certainly , we can find no evidence for the prominence feature on the 3rd night .
this is despite the fact that the emission should be well separated from the stellar line profile after orbital phase 3.645 .
this supports the idea that the prominence material is not long - lived , only lasting a couple of days before either draining back to the stellar surface or being ejected .
@xcite found that , while some prominences on speedy mic were still visible after 5 nights , others formed or disappeared over the course of one night . finally , unlike the prominences seen on speedy mic and ab dor which lie typically between 2 to 9 stellar radii from the stellar rotation axis ( with a concentration at the co - rotation radius ) , the feature observed in bv cen is far closer to the stellar surface .
if the prominence is located above the centre - of - mass of the binary , then this places it just 1.5 r@xmath22 ( assuming the roche - lobe volume radius for bv cen ) from the rotation axis of the secondary star .
although this is at odds with most h@xmath14 observations which show prominences at or beyond the co - rotation radius , clouds substantially closer to the stellar surface have been reported in hk aqr ( @xcite reported prominences 0.34 3.2 r@xmath22 above the stellar surface ) and re j1816 + 542 ( heights as low as 0.88 r@xmath22 , @xcite ) .
the height of the prominence feature seen in bv cen also agrees with x - ray observations of rapidly - rotating isolated stars , even when h@xmath14 observations of prominences of the stars in question reveal high cloud heights .
for instance , in chandra x - ray observations of ab dor , @xcite found that a significant fraction of the emission arose from compact regions near the stellar surface with heights of less than 0.3 r@xmath22 , and that the emitting corona does not extend more than 0.75 @xmath23 above the stellar surface .
obviously , the formation mechanism of these prominences and active regions necessitates more work . in particular ,
the apparent preference for ` slingshot prominences ' observed in cvs to form at low - velocity sites requires satisfactory explanation .
we have shown that the donor star in bv cen is highly spotted , the second cv donor for which this has been shown to be the case .
[ fig : plan ] highlights the striking similarities in spot distribution between bv cen and ae aqr , despite their quite different binary parameters and spectral types .
although bv cen has a higher spot coverage and more low - latitude spots than ae aqr , both show a high latitude spot displaced towards the trailing hemisphere . it would be interesting if all donor stars in cvs showed high latitude spots displaced in the same direction .
such a deflection could be the result of the orbital motion of the binary .
since the rotation axis lies outside the donor stars in these binaries , one may expect the play of coriolis ( and/or centrifugal ) forces on magnetic flux tube emergence to be different from single stars .
further observations , however , are required before we can definitely say that such a deflection of high - latitude spots takes place in cv donors .
another striking similarity between the two tomograms presented in fig .
[ fig : plan ] is the apparent chain of spots extending down from the polar regions to the @xmath1 point .
this is more evident in the tomogram of ae aqr , but probably only because there are fewer low latitude features which allows these spots to stand - out more on ae aqr .
we believe that this chain of spots is probably due to the impact of tidal forces due to the close proximity of a compact companion , and such a ` sub - white dwarf ' concentration of spots is also seen on the pre - cv v471 tau ( @xcite ) .
a concentration of starspots on the inner face of the donor stars in cvs may also have consequences for the accretion dynamics of these objects .
it has long been thought that starspots may be able to quench mass - transfer from the donor as they pass across the @xmath1 point , leading to the low - states seen in many cvs ( e.g. @xcite , @xcite ) .
indeed , in their study of the mass - transfer history of am her , @xcite concluded that such a model would require an unusually high spot - coverage near the @xmath1 point , or otherwise some mechanism that drives spots towards the @xmath1 point .
certainly , both ae aqr and bv cen do seem to show increased spot coverages towards the @xmath1 point in support of their conclusions .
the fact that we see more active regions near the @xmath1 point may also explain why we appear to see a preference for ` slingshot prominences ' to form above the donors inner face ( @xcite ; @xcite ; @xcite ; @xcite and again in this work ) .
certainly , the fact that prominence material at these locations will be illuminated by the accretion regions also means that observations will be biased towards detecting prominences in the region between the white dwarf and donor star in cvs .
furthermore , if surface magnetic fields are strong enough in the neighbourhood of the @xmath1 point this may also cause fragmentation of the mass flow , resulting in the inhomogeneous or ` blobby ' accretion seen in some cvs ( e.g. @xcite ; @xcite ) .
bv cen is the second cv donor star for which starspots have unambiguously been imaged on . again , as with ae aqr , we find a high ( 25 per cent ) spot coverage , and the high activity - level is further confirmed by the detection of a slingshot prominence .
comparison with the roche tomograms of ae aqr from @xcite show many similarities between the two systems despite the quite different fundamental parameters of these two binaries .
these spot distributions hint at the impact of tidal and/or coriolis forces on the emergence of magnetic flux tubes in these binaries , and suggest that the inner faces of cv donor stars are unusually heavily spotted . as always , further observations are recommended in order to confirm that such spot distributions are widely seen on cv donors .
if a fixed spot distribution were to be found , this would also suggest that differential rotation is suppressed ( or at least weak ) in cv donors , as suggested by @xcite .
caw is supported by a pparc postdoctoral fellowship .
ds acknowledges a smithsonian astrophysical observatory clay fellowship as well as support through nasa go grant nng06gc05 g .
ts acknowledges support from the spanish ministry of science and technology under the programme ramn y cajal .
the authors acknowledge the use of the computational facilities at sheffield provided by the starlink project , which is run by cclrc on behalf of pparc .
the starlink package photom was used in this work .
we would like to thank the observatories of the carnegie institution of washington for generously allowing us the use of the henrietta swope telescopes at las campanas observatory .
the 6.5 m landon clay ( magellan ii ) telescope at las campanas is operated by the magellan consortium consisting of the carnegie institution of washington , harvard university , mit , the university of michigan , and the university of arizona .
r. , shectman s. a. , gunnels s. m. , mochnacki s. , athey a. e. , 2003 , in iye m. , moorwood a. f. m. , eds , 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 .
1694 - 1704 ( 2003 ) .
mike : a double echelle spectrograph for the magellan telescopes at las campanas observatory . | we present roche tomograms of the g5g8 iv / v secondary star in the long - period cataclysmic variable bv cen reconstructed from mike echelle data taken on the magellan clay 6.5-m telescope .
the tomograms show the presence of a number of large , cool starspots on bv cen for the first time .
in particular , we find a large high - latitude spot which is deflected from the rotational axis in the same direction as seen on the k3k5 iv / v secondary star in the cataclysmic variable ae aqr .
bv cen also shows a similar relative paucity of spots at latitudes between 4050@xmath0 when compared with ae aqr .
furthermore , we find evidence for an increased spot coverage around longitudes facing the white dwarf which supports models invoking starspots at the @xmath1 point to explain the low - states observed in some cataclysmic variables . in total
, we estimate that some 25 per cent of the northern hemisphere of bv cen is spotted .
we also find evidence for a faint , narrow , transient emission line with characteristics reminiscent of the peculiar low - velocity emission features observed in some outbursting dwarf novae .
we interpret this feature as a slingshot prominence from the secondary star and derive a maximum source size of 75,000 km and a minimum altitude of 160,000 km above the orbital plane for the prominence . the _ entropy landscape _
technique was applied to determine the system parameters of bv cen .
we find @xmath2 = 1.18 @xmath3 m@xmath4 , @xmath5 = 1.05 @xmath6 m@xmath4 and an orbital inclination of @xmath7 = 53@xmath8 4@xmath0 at an optimal systemic velocity of @xmath9 = 22.3 km s@xmath10 .
finally , we also report on the previously unknown binarity of the g5iv star hd 220492 .
stars : novae , cataclysmic variables stars : spots stars : late - type stars : imaging stars : individual : bv cen techniques : spectroscopic |
there has been growing interest in the non - extensive statistical mechanics @xcite based on tsallis generalized entropy ( in @xmath5 unit ) : @xmath6 at least formally , tsallis entropy is an extension of conventional boltzmann - shannon ( bs ) entropy with one - real - parameter of @xmath1 . in the limit of @xmath7 , tsallis entropy eq .
( [ sq ] ) reduces to bs entropy , @xmath8 , since @xmath9 .
the parameter @xmath1 may be interpreted as a quantity characterizing the degree of non - extensivity of tsallis entropy through the so - called _ pseudo - additivity _ : @xmath10 where @xmath11 and @xmath12 denote two statistically independent sub - systems .
it is worth while to realize that the pseudo - additivity of @xmath13 is one of the crucial ingredients of tsallis non - extensive statistical mechanics .
in fact , the uniqueness of tsallis entropic form is proved @xcite for an entropy that fulfills the generalization of the shannon - kinchin axioms @xcite based on the pseudo - additive conditional entropy obeying the pseudo - additivity instead of additivity .
then what is a role of the pseudo - additivity ? by rewriting eq .
( [ sq ] ) as the following form , @xmath14 we see that the pseudo - additivity of tsallis entropy comes from the @xmath1-logarithmic function , which is defined by @xmath15 since it equips the pseudo - additivity as @xmath16 the inverse function of the @xmath1-logarithmic function is @xmath1-exponential one , which is defined by @xmath17^{\frac{1}{1-q}},\ ] ] for @xmath18 and otherwise @xmath19 .
as tsallis @xcite has already pointed out , the parameter @xmath1 plays a similar role as the light velocity @xmath20 in special relativity or planck s constant @xmath21 in quantum mechanics in the sense of a one - parameter extension of classical mechanics . unlike @xmath20 or @xmath21 ,
however , @xmath1 does not seem to be a universal constant .
thus it is a natural question whether @xmath1 is merely an adjustable parameter or not in the non - extensive statistical mechanics . in some cases the parameter @xmath1 has no physical meaning , but when it is used as an adjustable parameter the resulting distributions give excellent agreement with experimental data . in other but a few cases @xcite , @xmath1
is uniquely determined by the constraints of the problem and thereby @xmath1 may have a physical meaning .
recent studies of the characterization of mixing in one - dimensional ( 1d ) dissipative maps @xcite and of symbolic sequences @xcite seems to provide a positive answer to the above question since there exists the special value @xmath2 such that @xmath22 becomes linear .
for example , in the studies @xcite of mixing in simple logistic map , @xmath2 can be obtained by three different methods based on : i ) the upper bound of a power - law sensitivity @xmath23 to initial conditions ; ii ) the singularity indices in multi - fractal structure ; and iii ) the rate of information loss in terms of @xmath0 .
the remarkable fact is that all methods lead to the same value of @xmath24 , which may shed some light on the physical meaning of @xmath1 .
they established some connections among the sensitivity @xmath23 to initial conditions , tsallis entropy @xmath0 and the proper entropic index @xmath2 .
in particular we focus on the work of buiatti _
et al . _
@xcite , in which they have shown , for the symbolic sequences with length of @xmath25 , that the generalized block entropy @xcite is proportional to @xmath25 when the proper entropic index @xmath26 is used . in other words
, there may exist a proper entropic index @xmath26 , which may statistically characterize a non - extensive system .
in this work we study a reason why the generalized block entropy in the work of buiatti _
et al . _
@xcite is proportional to the length @xmath25 of symbolic sequences when we use the proper entropic index @xmath26 .
in particular we focus our attention on a role of the pseudo - additivity of the conditional entropy in characterizing a non - extensive system with the proper entropic index @xmath2 .
we reformulate the constraint of obtaining the proper entropic index @xmath26 as follows : the pseudo - additive conditional entropy becomes additive with respect to @xmath25 when the proper entropic index @xmath26 is used . in other words , for the special value @xmath2 of entropic index , the additivity of the conditional entropy is held in spite of that the involved subsystems are not statistically independent of each other .
the rest of the paper is organized as follows : in the next section we explain the constraint of obtaining the proper entropic index @xmath2 in the work @xcite of buiatti _
et al . _ , and propose our constraint of obtaining @xmath2 .
we then show the equivalence of the two constraints and discuss the underlying simple mechanism of why the conditional entropy becomes additive for the proper entropic index @xmath2 under the assumption of equi - probability . in section 3 , we discuss long - range correlation expressed by @xmath1-exponential function .
section 4 is devoted to our conclusions .
buiatti _ et al . _
@xcite showed , by studying a symbolic binary sequence @xmath27 with a long - range correlation , that for the probability @xmath28 of each path with length of @xmath25 , the generalized block entropy @xcite , @xmath29 is proportional to @xmath25 if and only if the variable index @xmath1 equals the proper entropic index @xmath2 .
we reformulate this constraint as the following .
the pseudo - additive conditional entropy @xcite , which is defined by @xmath30 should satisfy the _ additivity _
, @xmath31 when @xmath1 is equal to the proper entropic index @xmath26 . at first sight
our constraint seems to be paradoxical , since @xmath0 is pseudo - additive in general .
we explain in the followings that the equivalence of the two constraints which determine @xmath2 , and discuss an underlying simple mechanism connecting the variable entropic index @xmath1 with the proper index @xmath2 .
the method of buiatti _
et al . _
is rephrased as follows : @xmath32 is a linear function of @xmath25 when @xmath3 , i.e. @xmath33 where @xmath34 is a proportional constant , or equivalently @xmath35 subtracting @xmath36 from the both sides and after a little bit algebra , eq .
( [ linear - sq ] ) is rewritten as @xmath37 dividing the both sides of eq .
( [ linear - sq2 ] ) by @xmath38 and using the definition of the conditional entropy of eq .
( [ cond - ent ] ) , it is obvious that eq .
( [ linear - sq ] ) is equivalent to eq .
( [ additive - sq ] ) .
hence we have reformulated the method of buiatti _
_ as follows : the conditional entropy becomes additive when we use the proper entropic index @xmath2 .
having explained the equivalence of buiatti _ et al . _ and our methods of obtaining a proper entropic index @xmath2
, we now consider why the conditional entropy @xmath40 is proportional to the length @xmath25 of symbolic sequences when @xmath3 . under the assumption of equi - probability , tsallis entropy can be written in terms of the number of states @xmath41 for the symbolic sequences with the length of @xmath25 as @xmath42 then the conditional entropy of eq .
( [ cond - ent ] ) is expressed in terms of @xmath41 as @xmath43 suppose that the number of states @xmath41 obeys a power - law evolution , which can be well described by @xmath44 with the proper @xmath2 of a system of interest , where @xmath45 is another constant . the relation between @xmath45 and @xmath34 of eq .
( [ linear - sq ] ) is discussed in the next section .
substituting eq .
( [ w ] ) into eq .
( [ cond - sq ] ) , the corresponding conditional entropy can be written as @xmath46 = \ln_q [ \exp_{q^*}(l'_{q^ * } \ ; n ) ] .\ ] ] now we readily see that @xmath40 is proportional to @xmath25 , if and only if we set @xmath1 to @xmath2 . in other words , if @xmath47 obeys the @xmath2-exponential evolution of eq .
( [ w ] ) , then it is reasonable to use its inverse function in order to define the conditional entropy .
tsallis entropic description may be well suited for a long - range correlated system which obeys a power - law evolution described by @xmath1-exponential function .
then how can @xmath1-exponential function express long - range correlation ?
we here explain that long - range correlation may be expressed by the non - factorizability of @xmath1-exponential function into independent terms .
the long - range correlation in this case means the initial condition dependency of long duration .
it is known that @xmath1-exponential function can not be resolved into a product of independent terms unless @xmath48 .
for example @xmath49 is not resolved into the independent factors as @xmath50 .
instead it can be expressed as the product of the dependent factors as @xmath51 .
let us focus on the long - range correlation associated with the power - law evolution of @xmath41 described by eq .
( [ w ] ) . using eqs .
( [ lin - sq ] ) and ( [ equi - prob ] ) , the @xmath1-exponential dependency of @xmath52 can be expressed as @xmath53 = \exp_q [ s_q(1 ) + l_q
n ] \nonumber \\ & = & \exp_q[s_q(1 ) ] \cdot \exp_q[\frac{l_q n}{1+(1-q)s_q(1 ) } ] \nonumber \\ & = & \exp_q[s_q(1 ) ] \cdot \exp_q[\frac{l_q}{1+(1-q)s_q(1 ) } ] \nonumber \\ & \times & \cdots \times \exp_q [ \frac{l_q}{1 + ( 1-q)\{s_q(1)+n-1\ } } ] .
\label{long - range}\end{aligned}\ ] ] note that @xmath54 appears in all terms and this reflects the initial condition dependency of long duration .
this feature is consistent with the single - trajectory approach by montangero _
_ @xcite in which they fix a given initial condition in order to obtain the @xmath2 of the non - extensive version of kolmogorov - sinai entropy for the dynamics of the logistic map at the chaotic threshold .
because of the initial condition dependency of long duration , an averaging over many different initial conditions is not appropriate
. now let us focus on the relation between the proportional constants @xmath45 and @xmath34 in the previous section . from the second line of eq .
( [ long - range ] ) , we see that @xmath55.\ ] ] comparing this with eq .
( [ w ] ) , @xmath45 and @xmath34 are related by @xmath56 which is the same relation @xcite of @xmath57 between the lagrange multiplier @xmath58 of optimal lagrange multipliers ( olm ) method and that @xmath59 of tsallis - mendes - plastino one in canonical ensemble formalism , where @xmath60 denotes partition function .
we have proposed a constraint of obtaining the proper tsallis entropic index @xmath2 in describing the evolutions of correlated symbolic sequences with length @xmath25 .
the proper entropic index @xmath26 can be determined by requiring that the conditional entropy @xmath40 should be proportional to @xmath25 if and only if @xmath1 equals the proper entropic index @xmath2 . in other words
@xmath0 becomes _ additive _ for the proper @xmath2 .
it is the non - factorizability of @xmath1-exponential function into independent terms that can express a long - range correlation .
one of the authors ( t. w ) acknowledges s. abe , t. arimitsu and n. arimitsu for useful comments and valuable discussion at the 9th symposium on non - equilibrium statistical physics held at tsukuba , japan . c. tsallis , cond - mat/0010150 .
r. j. v. dos santos , j. math .
* 38 * ( 1997 ) 4104 ; k. s. fa , j. phys .
a * 31 * ( 1998 ) 8159 ; s. abe , phys . lett .
a * 271 * ( 2001 ) 74 . c. e. shannon and w. weaver , _ the mathematical theory of communication _ ( university of illinois pres , urbana , 1963 ) ; a. i. khinchin , _ mathematical foundations of information theory _
( dover , new york , 1957 ) . u. m. s. costa , m. l. lyra , a. r. plastino and c. tsallis , phys .
e * 56 * ( 1997 ) 245 .
v. latora , a. rapisarda , c. tsallis and m. baranger , phys .
lett . a * 273 * ( 2000 ) 97 .
f. a. b. f. de moura , u. tirnakli and m. l. lyra , cond - mat/0002163 .
m. buiatti , p. grigoline and l. palatella , physica a * 268 * ( 1999 ) 214 .
t. arimitsu and n. arimitsu , phys .
e * 61 * , ( 2000 ) 3273 ; preprint ( 2001 ) .
s. abe and a. k. rajagopal , physica a * 289 * ( 2001 ) 157 ; quant - ph/0003145 . m. ignaccolo and p. grigolini , cond - mat/0004155
. s. martnez , f. nicols , f. pennini and a. plastino , physica a * 286 * ( 2000 ) 489 . | tsallis non - extensive entropy @xmath0 enables us to treat both a power and exponential evolutions of underlying microscopic dynamics on equal footing by adjusting the variable entropic index @xmath1 to proper one @xmath2 .
we propose an alternative constraint of obtaining the proper entropic index @xmath2 that the non - additive conditional entropy becomes additive if and only if @xmath3 in spite of that the associated system can not be decomposed into statistically independent subsystems .
long - range ( time ) correlation expressed by @xmath1-exponential function is discussed based on the nature that @xmath1-exponential function can not be factorized into independent factors when @xmath4 . and non - extensivity , tsallis entropy , pseudo - additivity , power law 05.20.-y , 05.90.+m , 05.45.-a |
gravitational lensing has become an important tool in recent years for studying many aspects of the universe .
as it probes total mass , independent of the light distribution , it opens new windows onto studies of large scale structures where dark matter has dynamical importance . with new galaxy lenses being discovered each year , the statistical use of systems to study dark matter halo shapes and orientations is fast becoming possible .
most lensing systems lie at high redshift as this is where the lensing cross - section is optimal , however for detailed studies of the lensing galaxies themselves , lower redshift systems are better laboratories .
2237 + 0305 was discovered as a low redshift quadruple lens by huchra et al .
( 1985 ) . at @xmath0=0.039
it is a visibly extended early - type spiral galaxy with multiple luminous stellar components .
in addition , its proximity to us places its four lensed quasar images close to the centre of the galaxy at @xmath1700pc ( most lens images sit @xmath15kpc from the deflector s centre ) making the probed region of the galaxy more compact .
these features of the 2237 + 0305 system make it an ideal laboratory for studying the inner structure of the dark matter halo and properties of the lensing galaxy .
most galaxy lenses are early - type and studies of these systems can also provide important constraints on the dark matter ( for example treu , this meeting ) .
studies with spiral lenses are less frequent due to the fewer number known ( elliptical galaxies present a larger lensing cross - section probability ) but can provide information about the luminous components of the galaxy ( for example the system b1600 + 434 , maller et al .
2000 ) .
the importance of the stellar disk to the dynamical support of the galaxy is a controversial question in modern galactic astronomy .
sackett ( 1997 ) defined a maximal disk as one which provides 75 - 95 per cent of the galaxy s dynamical support . whether a disk is maximal or sub - maximal
has an impact on our understanding of disk dynamics and is therefore an important quantity to determine .
similarly , the mass - to - light ratios in the stellar components of a late - type galaxy are not well determined and their values can place constraints on star formation models .
traditional studies of spiral galaxies have used light profiles of the luminous components and spectroscopic dynamical information to determine these properties of the galaxy . in a complex system with multiple mass components
, a degeneracy exists between the mass - to - light ratio of the major stellar component and the mass scaling of the dark matter halo .
the additional information provided by lensing can break this degeneracy .
2237 + 0305 has been studied extensively in the twenty years since its discovery ( for example yee 1988 , foltz et al .
1992 , schmidt 1996 ) .
many simplified models of the galaxy s mass distribution have been published which focus on different aspects of the system .
we aim to model the major mass components with physically reasonable profiles and use these models to constrain the mass - to - light ratios , disk maximality and dark matter halo inner slope .
schmidt ( 1996 ) was the first to investigate the effect of the galaxy s bar on the lensed image positions .
he found the bar provides a significant torque to their positions and is thus a critical component to any lens model .
in addition to the bar , we model an inclined stellar disk , an elliptical bulge at the same position angle , and a dark matter halo . the disk is modelled as a thin structure with an exponential surface density profile , the bulge with a de vaucouleurs profile and the bar as a ferrers ellipse , in line with previous works .
the dark matter halo we model with both a flattened nfw profile ( navarro , frenk & white 1996 ) and as a softened isothermal sphere ( sis ) .
the parameters characteristic to these models have been taken from previous observational studies and are presented in table 1 . .fixed and free parameters and their source for the mass models of the four major components .
the label ` free ' refers to free parameters in the model and @xmath2 and @xmath3 denote the core radius and scale length respectively .
@xmath4schmidt 1996 , @xmath5yee 1988 , @xmath6racine 1991 , @xmath7irwin et al . 1989 .
[ cols= " < , < , < , < " , ] it is evident from the similarity of the mass - to - light ratios and the small nfw flattening that these two halo profiles would contribute comparable convergence and shear to the lensing .
figure 1 shows the rotation curve of a galaxy with the best - fitting parameters for the softened isothermal sphere halo and luminous components .
the bulge component clearly dominates the centre of the galaxy , providing the majority of the dynamical support within the image region . in the outer regions of the galaxy the dark matter component dominates , as expected from results of previous work deconvolving galactic rotation curves .
the best - fitting rotation curve using the flattened nfw dark matter halo is essentially the same .
the disk is clearly sub - maximal according to the definition of sackett ( 1997 ) , however the importance of the bulge in this system should not be forgotten .
this early - type spiral is not the ideal system for studying the importance of disk stellar support as the bulge clearly dominates the disk .
although the results presented are acceptable fits to the data , they are not likely to describe the real situation in the galaxy .
firstly , the flux ratios of the images have not been included in the minimisation procedure , even though they are additional constraints .
the first reason for this is the uncertain nature of ratios measured for most lensing systems .
q2237 + 0305 is not a radio - loud quasar and radio measurements are difficult .
optical fluxes probe a small region in the source plane and are subject to possible microlensing leading to spurious values .
agol , jones & blaes ( 2000 ) measured mid - ir fluxes for this system arguing the larger extent of the emitting region would reduce the contamination of microlensing and the waveband would minimise the effects of dust reddening .
we will use these measurements to compare with our best - fitting model .
the second reason for omitting the flux ratios in the analysis is the contention over their applicability when dark matter substructure is taken into account ( schechter & wambsganss , this meeting ) .
our dark matter is modelled as a smooth component .
comparison of the flux ratios shows excess flux in one of the images with both profiles for the dark matter .
this result is quite robust and suggests an underlying problem in the mass distribution .
in addition to this problem , the bulge mass - to - light ratios are extremely low compared with other studies of the bulge component in i - band ( see for example fukugita , hogan & peebles 1998 ) which place the figure closer to m / l=5.0@xmath8 . clearly , increasing the bulge mass changes the convergence and the shear in the image region due to its dominance of the mass there . without changing the remainder of the mass components , the image positions would be torqued out of place .
the location of the images defines the projected mass within that region quite accurately ( wambsganss & paczynski 1994 measured m@xmath9 = ( [email protected] ) @xmath11 10@xmath12h@xmath13 ) and therefore the bulge mass - to - light ratio can only be increased so far before it alone provides all of the convergence .
in addition , as the bulge is an elliptical component , increasing its contribution will increase the shear .
the one completely unknown component of the mass is the dark matter halo .
the best - fitting disk and bar sizes are what one would expect .
if the bulge is too small in our models then it would appear the dark matter distribution would need to be changed to accommodate this . the most likely outcome is a halo with little contribution within the image region but sufficiently flattened as to counteract the increased torque from the bulge . without further information , this is purely speculation .
additional dynamical information can help to answer these remaining questions . with only rotation data for the outer galaxy , critial regions of the rotation curve are unknown .
well measured points in the galactic centre will tightly constrain the allowed dynamics in the important image region , as shown in figure 2 .
it compares the expected observed rotation curve for this galaxy with a maximal bulge and the two different halo profiles .
this solution does not fit the image positions , but shows the possible difference in the inner rotation curve .
it is this splitting of curves that we hope to probe with future spectroscopy of the galaxy .
we have modelled the galaxy 2237 + 0305 with multiple mass components to investigate the mass - to - light ratios of the luminous components , the dynamical importance of the disk and bulge and the shape of the dark matter halo .
we require additional dynamical information to define the dark matter contribution accurately .
this system is potentially useful for constraining the mass - to - light ratio of the bulge component and the shape and inner slope of the dark matter halo .
agol , e. , jones , b. & blaes , o. 2000 , , 545 , 657 barnes , d. , webster , r. , schmidt , r. & hughes , a. 1999 , , 309 , 641 blanton , m. , turner , e. & wambsganss , j. 1998 , , 298 , 1223 foltz , c. , hewett , p. , webster , r. & lewis , g. 1992 , , 386 , l43 fukugita , m. , hogan , c. & peebles , j. 1998 , , 503 , 518 huchra , j. , gorenstein , m. , kent , s. , shapiro , i. , smith , g. , horine , e. & perley , r. 1985 , , 90 , 691 irwin , m. , webster , r. , hewett , p. , corrigan , r. & jedrzejewski , r. 1989 , , 98 , 1989 maller , a. , simard , l. , guhathakurta , p. , hjorth , j. , jaunsen , a. , flores , r. & primack , j. 2000 , , 533 , 194 navarro , j. , frenk , c. & white , s. 1996 , , 462 , 563 racine , r. 1991 , , 102 , 454 sackett , p. 1997
, , 483 , 103 schmidt , r. 1996 , msc thesis , univ . of melbourne yee , h. k. c. 1988 , , 95 , 1331 | we have studied the mass distribution in the lensing galaxy 2237 + 0305 using constraints from both gravitational lensing and photometric and spectroscopic observations .
we find that with sufficient dynamical information we can constrain the mass - to - light ratio of the luminous components and determine the dynamical contribution of the disk .
in addition , future observations should allow us to place constraints on the shape of the dark matter halo and its inner slope .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
the cosmological constant problem is usually described as a huge difference between the vacuum energy density derived by quantum field theories and the observed value of the cosmological constant .
the energy density associated to the vacuum fluctuations of free fields is a divergent quantity , and any natural cutoff we use to regularize it ( for example , the planck mass or the energy scale of the quantum chromodynamic ( qcd ) phase transition , the latest cosmological vacuum transition ) leads to values many orders of magnitude above the observed @xmath0 .
but the applicability of flat - spacetime quantum field theory in the expanding universe is doubtful , since a non - vanishing bare cosmological constant is not compatible with a flat spacetime .
the vacuum contribution to ( semi - classical ) gravity is only properly calculated when one considers a non - flat background and renormalizes the divergent contributions in some appropriate way .
this is a difficult task and , in general , may depend on the renormalization method used . in the simpler case of free conformal fields in a de sitter spacetime @xcite ,
the renormalized vacuum density is proportional to @xmath1 , where @xmath2 is the hubble parameter . in the high - energy limit
this result can be used to obtain a non - singular cosmology , with an initial quasi - de sitter phase , giving origin to a subsequent radiation dominate period @xcite . however , in the present universe the above result leads to a very tiny vacuum term . on the other hand
, it may well be possible that the observed cosmological term is just a cosmological constant of purely geometric nature . in this case
there would be no natural ( micro- ) physical scale associated to it .
furthermore , both the gravitational action of the quantum vacuum and the approximate coincidence between such a constant and the present matter density ( the cosmic coincidence problem ) would remain unexplained . as long as this issue
can not be solved on a fundamental level , ( semi- ) phenomenological models which avoid the theoretical difficulties pointed out above may help to get further insight into the problem . in this letter
we argue that , by modeling the cosmological `` constant '' in terms of a decaying vacuum energy , it is possible to obtain a model which is competitive with the standard one from the observational point of view . a crucial point here is to consider the contribution of interacting fields to the vacuum energy .
indeed , some authors have been arguing that the energy density associated to the qcd vacuum condensate in a low - energy de sitter ( or approximately de sitter ) spacetime is given by @xmath3 where @xmath4 mev is the energy scale of the qcd phase transition @xcite . in the present universe
this density is dominant when compared to the contribution of free fields , @xmath1 , and has the correct order of magnitude of the observed cosmological term . in this paper
we will trace back this scaling law @xmath5 to a process of matter production in the expanding spacetime .
let us consider a low - energy , spatially flat friedmann - lematre - robertson - walker ( flrw ) spacetime , where non - relativistic dark particles of mass @xmath6 are created out of the vacuum at a constant rate @xmath7 .
the particle number balance ( sometimes called boltzman equation ) for this process will be given by @xmath8 where @xmath9 is the particle number density at a given time .
it can be rewritten as @xmath10 where @xmath11 is the matter density .
this equation is characteristic of models with interaction in the dark sector @xcite .
this particle production from vacuum fluctuations necessarily involves a gravitational backreaction .
in addition to equation ( [ conservacao ] ) we have the friedmann equation @xmath12 where @xmath0 is the vacuum contribution to the energy density .
we shall assume @xmath13 for the vacuum pressure , admitting , however , @xmath0 to be time varying . the total energy density satisfies the conservation equation @xmath14 provided we take @xmath15 , where @xmath16 is a constant of integration , which corresponds to a purely geometric cosmological constant .
since there is no natural ( microphysical ) energy scale associated to this constant , it is disregarded here . in this way we have @xmath17 and equation ( [ conservacao2 ] )
can be rewritten as @xmath18 we can see that the particle creation is concomitant with a decay in the vacuum energy density . equation ( [ linear ] ) is the same time variation law ( [ schutz ] ) obtained in a quasi - de sitter spacetime when the contribution of the qcd condensate is taken into account @xcite . in this case
we have @xmath19 , where @xmath20 is the energy scale of the qcd phase transition . ] .
as discussed in @xcite , the association of the qcd energy scale with the observed vacuum energy is natural and , in some way , expected , since the qcd confinement , with breaking of chiral symmetry , is the last cosmological vacuum transition we know .
the qcd condensate formed after this low energy transition gives , in a de sitter ( or approximately de sitter ) spacetime , the correct value of the cosmological term ( see below ) .
furthermore , owing to the maximal symmetry of the de sitter background , the condensate pressure is the negative of its energy density , i.e. , the same equation of state associated to a cosmological constant .
dividing equation ( [ linear ] ) by @xmath21 , using @xmath22 and specifying to the present time ( subscripts @xmath23 ) , we obtain @xmath24 in a de sitter universe ( @xmath25 ) we would have @xmath26 .
that is , the particle creation rate would be equal ( apart from a numerical factor ) to the gibbons - hawking temperature associated to the event horizon @xcite . on the other hand , by taking @xmath27 for our present universe , we have @xmath28 and @xmath29 . the former relation is an expression of the old eddington - dirac large number coincidence @xcite , while the latter ( known as zeldovich s relation @xcite ) gives the correct order of magnitude for @xmath0 .
the exact value of @xmath7 in ( [ rate ] ) ( and hence of @xmath30 ) will be given by observations in the next sections .
once @xmath7 is fixed , no further fine - tuning will be necessary .
the cosmological model to be built will have the same number of free parameters as the @xmath0cdm model , namely @xmath31 and @xmath32 .
from this section on we will analyze the cosmological model arising from a vacuum energy density scaling with time in accordance with the ansatz discussed above .
when compared with the most precise observational constraints , the model will show a very good concordance and , in some cases , less tensions than in the standard case .
therefore , besides providing a possible physical basis for the cosmological term , in the lines of the previous sections , it will also show to be competitive from the observational point of view . with @xmath33
we obtain , from the friedmann equations , the solution @xcite @xmath34 ^ 2 + \omega_{r0 } ( 1+z)^4 \right\}^{1/2}\;,\ ] ] where @xmath31 is the present relative matter density , and we have added conserved radiation with present density parameter @xmath35 . as discussed in @xcite , for non - zero @xmath35 the expression ( [ hgeral ] ) is an approximate solution , differing only @xmath36 from the exact one , since @xmath37 . for @xmath38 ,
the solution ( [ hgeral ] ) is exact . for early times
we obtain @xmath39 , and the radiation era is indistinguishable from the standard one . on the other hand , for high redshifts the matter density scales as @xmath40 .
the extra factor @xmath31 - as compared to the @xmath0cdm model - is owing to the late - time process of matter production . in order to have nowadays the same amount of matter
, we need less matter in the past . or , in other words , if we have the same amount of matter in the past ( say , at the time of matter - radiation equality ) , this will lead to more matter today .
we can also see from ( [ hgeral ] ) that , in the asymptotic limit @xmath41 , the solution tends to the de sitter solution .
note that , like the @xmath0cdm model , the above model has only two free parameters , namely @xmath31 and @xmath32 . on the other hand
, it can not be reduced to the @xmath0cdm case except for @xmath42 . in this sense , it is falsifiable , that is , it may be ruled out by observations .
the hubble function ( [ hgeral ] ) is all we need to test the model against observations of snia and bao @xcite . however , when testing the position of the first peak of the cmb anisotropy spectrum , some care is needed .
the relation between the observed position of the peak and the acoustic scale @xmath43 is given by @xcite @xmath44 with @xmath45 evaluated at the redshift of last scattering @xmath46 .
the acoustic scale is defined , as usual , by @xmath47 with the sound velocity @xmath48 where @xmath49 and @xmath50 stand for the present density parameters of baryons and photons , respectively .
the relation ( [ l1 ] ) does not depend on the dark energy model , but only on pre - recombination physics .
however , the ratio @xmath51 is , in our case , given by @xmath52 with the extra factor @xmath31 as compared to the @xmath0cdm expression .
note that , in all the above expressions , there is a dependence on @xmath32 through the relative densities of baryons , photons and radiation , as well as through the redshift of last scattering . in our analysis
we will fix these parameters ( @xmath53 , @xmath35 , @xmath46 ) to the concordance values of the standard model .
the only free parameters to be adjusted will then be @xmath31 and @xmath32 .
the result of a joint analysis of snia , bao and the position of the first peak of the cmb spectrum is shown in figure [ figsdss ] . in the left panel
we show the superposition of the corresponding confidence regions in the ( @xmath54,@xmath31 ) plane , where @xmath55(km / s)/mpc is the dimensionless hubble constant .
the central panel shows the same for the @xmath0cdm model . in the right panel a comparison between the two models is made , after marginalizing over @xmath54 . in this analysis
we have used the sdss snia compilation with the mlcs2k2 light - curve fitter @xcite .
other current compilations make use of salt or salt2 fitters which , however , are not model independent . a more complete analysis , including the union2 and constitution compilations , can be found in @xcite . in the case of the constitution sample calibrated with mlcs2k2 ,
the results are very similar to those presented here .
note that for the standard model the relative matter density is higher than the usually accepted concordance value ( @xmath56 ) .
this is a characteristic of the mlcs2k2 fitter , also present when we use it to calibrate the constitution sample .
the analysis of the matter power spectrum was performed in ref .
@xcite , where , for simplicity , baryons were not included and the cosmological term was not perturbed . in a subsequent publication a gauge - invariant analysis , explicitly considering the presence of late - time non - adiabatic perturbations , has shown that the vacuum perturbations are indeed negligible , except for scales near the horizon @xcite .
therefore , the result of ref .
@xcite can be considered a good approximation .
an updated analysis is shown in the left panel of figure [ figpowersp ] , together with the 2dfgrs data points @xcite and , for comparison , the concordance @xmath0cdm fitting as provided by the bbks transfer function @xcite .
it is also shown the @xmath57 confidence region for our fitting .
rcccc + & & + & & & & + + union2 ( salt2) ...... &@xmath58 & 1.063 & @xmath59 & 1.027 + sdss ( mlcs2k2 ) ....... & @xmath60 & 0.842 & @xmath61 & 1.231 + constitution ( mlcs2k2[17 ] ) ....... & @xmath62 & 1.057 & @xmath63 & 1.384 + in the right panel of figure [ figpowersp ] we show the superposition of the @xmath64 and @xmath57 confidence regions for the four tests we are considering here ( sne ia , bao , first peak of cmb and lss ) , as well as the concordance ellipses of our joint analysis .
the concordance value for the matter density is @xmath65 ( @xmath57 ) .
the hubble parameter is smaller than in other estimations , but it has the same value obtained when the same snia sample is analyzed for the @xmath0cdm model ( see fig . [ figsdss ] ) . in what concerns the matter density parameter ,
it is about @xmath66 higher than the flat standard model value for the sdss ( mlcs2k2 ) compilation ( @xmath67 ) . in both models
the obtained values of @xmath31 are higher than the usually accepted concordance value for the flat @xmath0cdm model , and the same occurs when one uses the constitution sample calibrated with mlcs2k2 . a discussion about the present tension between different snia samples and fitters , in both models , can be found in @xcite . here
we show , in table i , the best - fit results for @xmath31 ( with @xmath54 marginalized ) with three samples of supernovas : the sdss and constitution calibrated with the mlcs2k2 fitter , and the union2 ( which is calibrated with salt2 ) . in figure
[ figure03 ] , on the other hand , the reader can find the evolution of the scale factor ( left panel ) and of the relative energy densities of radiation , matter and dark energy ( central and right panels ) in the two models .
the concordance value obtained for @xmath31 is in agreement with estimates based on peculiar velocity measurements for galaxy pairs @xcite , but only in marginal agreement with dynamical estimations from clusters @xcite .
it is important to emphasize , however , that such estimations are not very precise and that they involve a different scale .
note also that the present result has an approximative character .
as discussed in @xcite , the inclusion of baryons in the analysis of the power spectrum may lead to a difference of about @xmath66 in the estimation of @xmath31 . with the concordance values of @xmath54 and @xmath31 in hand ,
we can obtain the age of the universe , as well as the redshift of transition between the decelerated and accelerated phases .
they are given , respectively , by @xcite @xmath68 and @xmath69^{2/3}-1.\ ] ] this leads to @xmath70 and @xmath71 , in good agreement with standard predictions and astronomical limits @xcite .
for @xmath72 we have @xmath73 gyr .
it is not easy to perform a full analysis of the cmb anisotropy spectrum with alternative models , specially if it involves new processes like particle production , for example . nevertheless , for at least some of the features we do not expect significant changes .
first of all , radiation and baryons are assumed to be separately conserved ( there is no photon fluctuations from the vacuum at the tree level , and baryons are too heavy to be produced at late - times ) .
therefore , the cmb temperature , the acoustic horizon at the last scattering and the relative height of the peaks will not be affected .
on the other hand , the redshift of equality between matter and radiation in our model is given by @xmath74 where we can note again an extra factor @xmath31 as compared to the standard @xmath0cdm result . by using our concordance matter density @xmath75
, we obtain @xmath76 , which does not differ too much from the standard value .
( this is simply because @xmath77 .
therefore , a relative matter density of around @xmath78 in the present model leads to the same @xmath79 of the standard model with @xmath80 ) . for this same reason
, the matter density at the time of equality will be approximately the same in both models .
[ by the way , this result for @xmath79 guarantees that the turnover of the matter power spectrum is correctly placed ( see fig .
[ figpowersp ] and @xcite ) ] .
as the redshift of last scattering is exactly the same @xcite , and the recombination does not depend on a particular dark energy model , both last scattering and recombination will occur in the matter dominated phase .
all these results indicate that the absolute height of the peaks will not change either .
( see the left panel of their figure 3 ) . on the other hand
, they claim that this density leads to a matter power spectrum in disagreement with lss data , contrary to what we have obtained in @xcite . our present results , however , agree with those obtained previously in reference @xcite . ]
the only important consequence of matter production on the cmb spectrum could be owing to the integrated sachs - wolf effect ( isw ) . in the right panel of figure [ figdcontrast ]
we plot the gravitational potential @xmath81 ( @xmath82 is the matter density contrast and @xmath83 is the comoving wavenumber ) as a function of the redshift , for the einstein - de sitter , @xmath0cdm and @xmath84cdm models , for @xmath85mpc .
the isw effect is proportional to the integrated change of the gravitational potential . as the right panel of fig .
[ figdcontrast ] shows , the slope of the potential of our model differs from the corresponding slope of the @xmath0cdm model .
but also here one may expect this effect to be small . in the @xmath0cdm model the potential decreases by @xmath86% between the last scattering and the present time , in our @xmath84 model it decreases by @xmath87% .
note , however , that fig .
[ figdcontrast ] is preliminary since it is based on a newtonian analysis @xcite . here
, a more quantitative , fully relativistic analysis is necessary . on the other hand , as can also be seen in fig .
[ figdcontrast ] , there is a small difference in the gravitational potential as compared to the standard model , a difference that will affect the normalization of the cmb spectrum , since the sachs - wolf effect is proportional to @xmath88 .
but the gravitational potential is proportional to the matter density contrast as well .
therefore , if the latter is correctly normalized ( by using , for example , the bbks transfer function @xcite , as was done , e.g. , in @xcite and in the left panel of fig .
[ figpowersp ] ) , the cmb spectrum will also have the correct normalization .
in other words , in order to have the correct normalization today , we have more power in the cmb spectrum at high redshifts as compared to the standard case .
the coincidence problem within the @xmath0cdm model is usually phrased as : why @xmath89 and @xmath0 are of the same order of magnitude just at the present epoch ?
our model relies on dark - matter particle production at a constant rate . from the outset
, this rate is a free parameter . by eq .
( [ rate ] ) it is assumed here to be of the order of the hubble constant . for our model the coincidence problem translates into : why the rate @xmath7 is of the order of the present - time hubble rate @xmath90 ( and not of the order of the hubble rate at a different time ) ?
any other value of @xmath7 would not be compatible with the observed cosmological dynamics . compared with the @xmath0cdm model , the coincidence problem
is alleviated , however , in the sense that @xmath91 @xcite , whereas in the @xmath0cdm model one has @xmath92 .
any decay of the ratio @xmath93 that is less than @xmath94 can be considered as an alleviation of the coincidence problem @xcite .
this difference in the background dynamics also influences the behavior of the perturbations . as the left panel of fig .
[ figdcontrast ] shows , the present model deviates from an einstein - de sitter universe already at higher redshifts than the @xmath0cdm model .
moreover , while the perturbations within the latter approach a constant value for decreasing @xmath95 , in our model there appears a maximum around the present epoch and perturbations are suppressed for @xmath41 .
the origin and nature of the dark sector is probably the central theoretical problem of modern cosmology .
the non - zero tiny value of the cosmological term and its coincidence with the matter density are some important faces of the question .
relating dark energy with vacuum fluctuations is an old difficulty , but one may expect a potential solution if this problem is appropriately considered in curved spacetime .
several models of dynamic vacuum have been suggested along these lines .
models for accelerated expansion based on matter creation in the expanding background have previously been proposed , e.g. , in @xcite .
although a relation between dynamical dark energy and particle production is not mandatory , we think it represents an interesting option .
this paper is based on the assumption that the late - time spacetime expansion can extract non - relativistic particles from the quantum vacuum . in the case of a constant creation rate ,
the backreaction of this process leads to a vacuum density linearly proportional to the hubble parameter , a result previously obtained from estimations of the gravitational contribution of the qcd vacuum condensate .
the resulting cosmological model is shown to be in accordance with the most precise observational tests , namely those which involve snia , the anisotropy spectrum of the cmb , the lss distribution and bao , albeit for higher values of the current matter density than in the standard model .
another difference to the @xmath0cdm model may be a modified isw effect .
this will be a subject of future research .
this work was partially supported by cnpq and faperj .
l. h. ford , phys . rev . * d11 * , 3370 ( 1975 ) ; j. s. dowker and r. critchley , phys . rev . *
d13 * , 3224 ( 1976 ) ; p. c. w. davies , phys . lett . *
b68 * , 402 ( 1977 ) ; a. a. starobinsky , phys . lett . *
b91 * , 99 ( 1980 ) .
r. schutzhold , phys .
lett . * 89 * , 081302 ( 2002 ) ; f. r. klinkhamer and g. e. volovik , phys .
d * 79 * , 063527 ( 2009 ) .
see also s. banerjee et al .
, phys . lett . *
b611 * , 27 ( 2005 ) , f. r. urban and a. r. zhitnitsky , phys . rev . * d80 * , 063001 ( 2009 ) ; phys . lett . *
b688 * , 9 ( 2009 ) ; nucl . phys . *
b835 * , 135 ( 2010 ) ; n. ohta , phys .
* b 695 * , 41 ( 2011 ) , r. g. cai _
_ , phys . rev . * d84 * , 123501 ( 2011 ) .
j. a. s. lima , l. r. w. abramo and a. s. m. germano , phys . rev . * d53 * , 4287 ( 1996 ) ; j. s. alcaniz and j. a. s. lima , astron
. astrophys . * 349 * , 729 ( 1999 ) ; w. zimdahl , phys .
d * 61 * , 083511 ( 2000 ) ; l. p. chimento , a. s. jakubi , d. pavon and w. zimdahl , phys . rev . * d67 * , 083513 ( 2003 ) ; j. a. s. lima , f. e. silva and r. c. santos , class .
. grav . * 25 * , 205006 ( 2008 ) ; g. steigman , r. c. santos and j. a. s. lima , jcap * 0906 * , 033 ( 2009 ) . | we demonstrate that creation of dark - matter particles at a constant rate implies the existence of a cosmological term that decays linearly with the hubble rate .
we discuss the cosmological model that arises in this context and test it against observations of the first acoustic peak in the cosmic microwave background ( cmb ) anisotropy spectrum , the hubble diagram for supernovas of type ia ( snia ) , the distance scale of baryonic acoustic oscillations ( bao ) and the distribution of large scale structures ( lss ) .
we show that a good concordance is obtained , albeit with a higher value of the present matter abundance than in the @xmath0cdm model .
we also comment on general features of the cmb anisotropy spectrum and on the cosmic coincidence problem . |
the karlsruhe tritium neutrino experiment ( katrin ) is a direct , model - independent search for the absolute mass of the electron antineutrino @xcite .
the highest energy electrons from the beta decay of molecular tritium ( @xmath2 ) will be selected by a mac - e spectrometer @xcite and tagged with a focal - plane detector ( fpd ) . the overall sensitivity to antineutrino mass depends critically on the minimization of backgrounds , which can be achieved by placing the fpd in extreme high vacuum ( xhv ) ( @xmath3pa ) and avoiding materials with high natural radioactivity near the fpd .
these conditions constrain the design of the fpd and preclude many standard detector construction materials and techniques .
therefore we have developed a novel scheme to support and read out the fpd .
-diode pixels .
dots denote points on the device that are contacted by pogo pins .
see text for details . ]
the katrin fpd ( fig .
[ fig : fpd ] ) is a monolithic silicon @xmath0-diode array manufactured by canberra , belgium .
the @xmath4-type silicon substrate wafer is 500@xmath1 m thick with a diameter of 114 mm .
it is bare , with no housing or backing of any kind .
the entrance , or front side , is uniformly @xmath5-doped with no segmentation .
the back side is @xmath6-doped in 148 individual 44.1-mm@xmath7 pixels arranged in a circular dartboard pattern surrounded by a continuous guard ring .
the @xmath8 crystal orientation is perpendicular to the surface .
the outermost pixels extend to a diameter of 90 mm .
a guard ring extends from a diameter of 90 mm out to 94 mm .
the pixels and guard ring are covered with titanium nitride ( tin ) for electrical contact to the front - end electronics .
the guard ring is to be held at signal reference potential to sink surface currents that might otherwise flow between the outer pixels and regions of the surface held at bias potential . outside the guard ring
there is a coating of tin ( but no doping ) from a diameter of 100 mm extending around the edge of the wafer to the @xmath5 doping on the front side .
the tin on the front side extends a few mm from the edge , but the front is otherwise unmetallized .
this configuration allows the bias potential on the front side to be applied via connections made on the back side adjacent to signal connections .
tin was chosen as a coating over more commonly used metals such as aluminum because it is non - oxidizing .
the resistance of thin oxide layers is likely to have prevented good electrical contact to the front - end electronics .
silver , though nonoxidizing , is excluded because it would create a background from an atomic fluorescence near in energy to the highest energy @xmath2 beta - decay electrons of interest to katrin .
the radiopurity requirements also limit the choice of techniques for connecting the detector to its front - end electronics .
in particular , wire bonding is avoided because of the radioactivity of ceramic substrates typically used in the process .
the process of bonding ceramic to silicon can also create localized regions of very high stress because of dissimilar coefficients of thermal expansion .
instead , the connection between detector and electronics is made by an array of spring - loaded pogo pins .
the pins are manufactured by interconnect devices , inc .
( item ss-30 with tip option j ) .
they are made of inherently radiopure and xhv - compatible materials .
the fpd is supported at a diameter of 99 mm on its front side while the pins press into the tin on the back side .
the front - end electronics are on the opposite side of a feedthrough in a region of high vacuum ( @xmath9pa ) . there will be a total of 184 pins pushing on the fpd ; one for each of the 148 pixels , 12 contacting the guard ring and 24 applying bias to the outermost ring .
these pins will exert a total force of up to 50n on the fpd , which will deform it , causing mechanical stress up to 50mpa in the silicon bulk .
mechanical stress reduces the band gap energy in silicon and thus increases the bulk - generated reverse - bias leakage current through a @xmath10 junction with respect to unstressed silicon @xcite . in this model , 50mpa of stress in the region of the junction causes approximately 25% increase in leakage current .
stress can also induce an increase of approximately 10% in junction capacitance for @xmath11 and @xmath12 junctions under a few volts of reverse bias @xcite .
both effects would tend to degrade the energy resolution of a silicon @xmath0-diode detector by several percent , if they apply to such devices .
most troubling for katrin is a report @xcite demonstrating that stress causes leakage current to increase by up to four orders of magnitude in a few specimens of outer silicon tracker sensors at the large hadron collider s compact muon solenoid ( cms ) .
no mechanism for this correlation is proposed , though bulk defects in the silicon are ruled out as the cause .
the behavior was observed only in a small subset of the production line cms devices from one of two manufacturers .
prototypes from the same manufacturer did not show the correlation .
the mechanisms described in @xcite can not account for such large increases .
order - of - magnitude leakage - current increases are predicted for stressed small - pitch devices @xcite , though it seems unlikely that that model would apply to some but not all of the cms devices .
it is not clear whether the expected mechanical stress would cause similar behavior in the katrin fpd . in this work ,
we demonstrate the efficacy of the katrin fpd connection scheme by implementing it on a prototype @xmath0-diode array similar to the fpd in dimensions and doping architecture .
properties of this detector are measured while unstressed and for a range of pogo - pin applied stresses up to a maximum of 3042mpa .
no stress - induced effects on the detector performance are observed .
-diode array used for these measurements , viewed from the pixelated side .
solid circles are tin - coated actively doped @xmath0-diode regions which are contacted by pogo pins .
open circles are tin - coated actively doped @xmath0-diode regions which are not contacted by pogo pins .
all pogo pins contact at the center of a diode region .
the numbers 015 appear just above the sixteen channels instrumented for these measurements .
see text for details . ]
canberra , belgium provided a prototype @xmath0-diode array for evaluation ( fig .
[ fig : caracmapbw ] ) .
the prototype s doping architecture is identical to that of the fpd except that the prototype has a thicker @xmath5 layer on the entrance side and its window is metalized with aluminum . like the fpd ,
the prototype array is monolithic on a 500-@xmath1m - thick silicon wafer with the @xmath8 crystal orientation perpendicular to the surface .
the back side is also coated with tin for ohmic contact to the pixels . the wafer measures 127 mm in diameter and has 158 8-mm@xmath7 pixels .
the apparatus constructed for these tests allows for measurements of the prototype s basic @xmath0-diode detector properties under controlled applied loads from a pogo - pin array .
the pogo - pin array is lowered from above .
the entrance side of the detector faces down towards a sealed 10-@xmath1ci @xmath13 source , placed 6.4 cm below the detector .
preamplifiers are mounted above the pogo - pin array .
the entire apparatus resides inside an electrically - shielded light - tight enclosure .
the detector is held firmly between two delrin rings with inner diameters of 110 mm .
a ring of electrically conductive elastomer sits between the lower ring and the detector in a recess cut into the delrin .
the recess is slightly shallower than the thickness of the uncompressed elastomer so that when the upper ring is attached the elastomer is compressed for good electrical contact to the detector .
the recess extends to the outer edge of the holder in eight channels .
positive bias potential is applied to the detector via wires pinched between the elastomer and the upper delrin ring in each of these channels . for control measurements of the detector s performance with no stress , an additional plug of delrin and elastomer
is inserted to support the detector from below the front side .
the apparatus is shown in fig .
[ fig : apparatus ] . electrical contact to each pixel is made by a spring - loaded conducting pogo pin .
the pins are manufactured by interconnect devices , inc .
( item number ss30 with a rounded tip [ option j ] and a stainless steel spring ) .
each pin is preloaded such that it takes a force of 0.137n to initiate any compression of the plunger and an additional force of 0.133n per 1000@xmath1 m of compression .
the pogo - pin array is aligned with the detector such that each pin contacts the center of one pixel .
close mechanical tolerances between the array and the detector holder preclude contacting 30 of the outermost pixels , and one interior pin was damaged during construction of the apparatus , leaving a total of 127 pogo pins applying force to the detector .
the pins are mounted and soldered into a stack of five 1.57-mm - thick fr4 circuit boards which is much more rigid than the detector .
the pins are positioned such that the variation in the relaxed vertical position of the tips is less than 25@xmath1 m .
a micrometer sets the pogo - pin array translation with an accuracy of 25@xmath1 m , up to a maximum of 1020@xmath1 m with respect to initial detector contact .
uniformity of the applied load is ensured through precise alignment of the pogo - pin array relative to the detector .
the height of all surfaces nominally parallel to the detector deviates less than 25@xmath1 m , even under the maximum loads allowed by the apparatus .
a blank silicon wafer with identical dimensions as the prototype detector was installed into the apparatus .
[ fig : deflections ] shows the displacement of the center of the blank wafer and the smaller deflection of the center of the pogo - pin array caused by the reaction force . due to these deflections
, care must be taken to ensure that all pins are compressed at least to the manufacturer s specification of 380@xmath1 m because electrical contact is inconsistent when the pins are under - compressed .
the array must be translated more than 760@xmath1 m past initial contact with the detector so that pins near the center of the detector will meet the minimum requirement .
as can be seen in fig .
[ fig : deflections ] , when the array is translated 760@xmath1 m , the pin compression in the center is reduced by the 340@xmath1 m deflection of the detector in addition to the 60@xmath1 m deflection of the array itself , resulting in only 360@xmath1 m compression of pins at the center .
calculation of the total force on the detector by the entire array is complicated by the deformation of the detector under the applied load .
the compression of each pin and deflection of the detector depend on the radial position in a way that is not exactly known .
the more relevant quantity is the maximum stress in the detector , which can be estimated without knowing the total load , provided that the displacement @xmath14 of the center of the detector is known ( fig .
[ fig : deflections ] ) .
mechanically , our detector is a circular plate with fixed edges .
the appropriate relations are @xcite : @xmath15 where @xmath16 is the maximum stress in the silicon , @xmath17 is the total load , @xmath18 is the thickness of the detector , @xmath19 is the radius of the supporting ring , and @xmath20 is the young s modulus of silicon . the load , @xmath17 ,
is trivially eliminated : @xmath21 the young s modulus of silicon ranges from 130185gpa , depending on the crystal orientation with respect to the direction of applied stress .
these formulae are appropriate for uniformly distributed loads and deflections of less than 50% of the plate thickness . however , the load on our detector has some position dependence due to the variation in pogo - pin compression caused by bowing of the detector , and our detector bows by 80% of its thickness under the maximum load allowed by the apparatus ( fig .
[ fig : deflections ] ) .
therefore we do not expect stresses calculated with equations ( [ eqn : stress1 ] ) and ( [ eqn : stress2 ] ) to be very precise , but take them to be good estimates of the lower limit with the correct order of magnitude
. with these caveats in mind , equation ( [ eqn : stress2 ] ) indicates that our detector will have a maximum stress of 3042mpa at the maximum array translation allowed by our apparatus .
the stated range is due to the unknown orientation of local stresses within the detector bulk .
the custom - made charge - sensitive preamplifiers are prototypes of those to be used for katrin . because of limited availability of preamplifiers , only eight pixels of the detector
are tested simultaneously .
a total of sixteen pixels were tested in two separate series of measurements .
they were chosen to give a representative sample across a diameter of the detector .
the stress is not uniform throughout the volume , but the circular symmetry of the holder guarantees that any stress depends only on the distance from the center of the detector . during each pass ,
all pixels neighboring the eight active channels were held at ground to ensure uniform electric fields in the active pixels and also to provide an alternative path for any surface currents which might flow on the detector .
the ground connection was temporarily removed to test for stress - induced surface currents .
no difference was observed .
all of the leakage current through a detector pixel passes through a 5% precision 500m@xmath22 resistor on its preamplifier , adding a dc offset to the output signal .
we compute the leakage current as the difference between the dc offsets of the output signal when the detector is biased to @xmath23 and when it is floating at @xmath24 with the bias supply physically disconnected , divided by the resistance : @xmath25/500\,{\rm m}\omega.\ ] ] this relation was validated to an accuracy of @xmath26% using a picoammeter in series with the single bias supply such that it read the total current drawn by all eight instrumented channels .
the picoammeter introduced a small amount of noise to the output signals , so it was removed during normal running conditions .
the bias potential is @xmath27v for all measurements reported here .
detector outputs are split into two branches by a custom - made frequency crossover circuit .
one branch carries components of the signal above a 3-db cutoff frequency of 68khz into a vme - based shaper / adc unit described in @xcite .
the dc offsets , @xmath28 , are carried on the other branch into an acromag ip320a adc housed in the same vme crate .
data are acquired from the vme crate by orca software @xcite running on a macintosh computer .
the setup is shown schematically in fig .
[ fig : signalflow ] .
the voltages on the dc lines are sampled once per second and these samples are then averaged during two - minute intervals . the measurements for each translation of the pogo - pin array consist of two such intervals one to establish @xmath29 for every channel before the detector is biased and another to establish @xmath30 for every channel once the detector is at full bias .
the statistical uncertainties in the leakage currents are due to the rms fluctuation of the sampled voltages .
the resistor precisions are not taken into account since they affect all measurements with a given preamplifier by the same proportion , and we are interested only in changes of the leakage currents .
an additional independent systematic uncertainty of @xmath31pa is added to account for the temperature - dependent gate leakage of the fet at the preamplifier input ( see below ) .
the temperature of the apparatus is monitored but it is not controlled .
the temperature inside the light - tight enclosure is measured with a thermistor mounted on the lower surface of the delrin holder but exposed to air .
the temperatures of the detector and electronics are not directly measured .
the temperature is recorded at the beginning and end of each two - minute dc - offset sampling period .
the nominal temperature for a single leakage current measurement is the average of four readings , two for each of two sampling periods .
the nominal temperature range for the measurements presented here is 24.425.6@xmath32c , though the four individual values which determine any particular nominal temperature never differ by more than 0.1@xmath32c .
we assume that fluctuations in the detector and electronics temperatures are the same size as the fluctuations read on the thermistor .
we therefore take the uncertainty in the nominal temperature for each measurement to be 0.1@xmath32c .
our results are demonstrated in two ways .
the first examines the pulse - height resolution of @xmath13 gamma - rays as the stress in the detector is increased .
the second method is a direct measurement of leakage currents as discussed above .
neither method reveals any effects attributable to the stress .
the results are qualitatively the same for all tested pixels . for brevity ,
we present results from pixel 8 only .
[ fig : spectra ] shows the raw @xmath13 gamma - ray pulse - height spectra obtained with seven different pogo - pin array translations from a typical pixel of our detector .
gamma rays from a single typical pixel for seven different translations of the pogo - pin array in steps of 130@xmath1 m increasing from the initial compression of 380@xmath1 m .
different colors represent different translations .
there is no statistically significant difference between these spectra . ]
the peak corresponding to the 59.5kev gamma is labelled for reference .
the recognizable gamma - ray spectrum of @xmath13 demonstrates the basic functionality of our mounting scheme
the detector works under stress and the pogo pins make good electrical contact to the tin - coated pixels .
[ fig : rawleakage ] shows the leakage current measurements from a typical pixel . ) . ]
the left plot shows the leakage current versus the array translation , which corresponds to increasing mechanical stress from left to right . the right plot shows the same leakage currents versus the temperature at the time they were measured .
leakage currents in diode detectors are thermal in nature so it is necessary to account for temperature dependence when searching for stress - induced changes .
the thermally generated leakage current @xmath33 is proportional to the rate of electron - hole pair creation and is thus related to the absolute temperature @xmath34 by @xcite @xmath35 where @xmath36 is the bandgap energy and @xmath37 is the boltzmann constant .
the normalization @xmath38 accounts for the device geometry and all factors contributing to charge - collection efficiency in the electronics ( recombination etc . ) .
the current measurements made with the detector unsupported are combined with those made with the detector supported and fit with a single function of the form ( [ eqn : ivst ] ) , leaving the normalization @xmath38 as the only free parameter .
the data fit well ( @xmath39 ) , indicating that all observed variation in the current measurements can likely be attributed to the temperature dependence of bulk - generated leakage currents .
we have demonstrated the efficacy of a new mounting scheme for monolithic silicon @xmath0-diode arrays .
the scheme consists of contacting tin coated pixels with an array of spring - loaded pins .
we observed no measurable effects of the associated mechanical stress on the performance of our monolithic silicon @xmath0-diode array for bulk stresses up to a few tens of mpa .
the large leakage current increases reported in reference @xcite were not observed in the sixteen tested pixels of our prototype detector .
all observed effects can be attributed to temperature variations .
the pogo - pin contact scheme to be used by katrin to connect its fpd to front - end electronics is not expected to have any problems related to the pin forces on the detector .
this research was supported by the us dept . of energy division of nuclear physics through
grant de - fg02 - 97er41020 .
the authors wish to thank marijke keters , mathieu morelle and the rest of their team at canberra in olen , belgium for technical support of this research and for the design and fabrication of katrin fpds .
99 j. angrik et al . , katrin design report , 2004 .
katrin / publications / documents / designreport2004 - 12jan2005.pdf .
v.m . lobashev and p.e .
spivak , nuclear instruments and methods a 240 ( 1985 ) 305 .
m. furuhashi and k. taniguchi , journal of applied physics 103 ( 2008 ) .
v. p. gopinath et al .
, ieee electron device letters 23 ( 2002 ) 312 .
a. j. furgeri et al .
, nuclear instruments and methods a 560 ( 2006 ) 108 .
p. smeys et al .
, ieee transactions on electron devices 46 ( 1999 ) 1245 .
e. oberg et al . , machinery s handbook , 27th ed .
, industrial press inc , 2004 .
amsbaugh et al . ,
nuclear instruments and methods a 579 ( 2007 ) 1054 . m.a .
howe et al . , ieee transactions on nuclear science 51 ( 2004 ) 878 .
knoll , radiation detection and measurement , 3rd ed .
, john wiley and sons , inc . , 2000 . | the karlsruhe tritium neutrino experiment ( katrin ) will detect tritium beta - decay electrons that pass through its electromagnetic spectrometer with a highly - segmented monolithic silicon @xmath0-diode focal - plane detector ( fpd ) .
this @xmath0-diode array will be on a single piece of 500-@xmath1m - thick silicon , with contact between titanium nitride ( tin ) coated detector pixels and front - end electronics made by spring - loaded pogo pins .
the pogo pins will exert a total force of up to 50n on the detector , deforming it and resulting in mechanical stress up to 50mpa in the silicon bulk .
we have evaluated a prototype @xmath0-diode array with a pogo - pin connection scheme similar to the katrin fpd .
we find that pogo pins make good electrical contact to tin and observe no effects on detector resolution or reverse - bias leakage current which can be attributed to mechanical stress . , , , , , , , , , silicon , @xmath0 diode , mechanical stress , leakage current , titanium nitride , pogo pin . 29.40.gx , 29.40.wk |
the classic black hole thermodynamics @xcite relates the mass @xmath1 , surface gravity @xmath2 , and outer horizon area @xmath3 of a black hole solution to the energy , temperature , and entropy ( @xmath4 , @xmath5 , and @xmath6 , resp . ) according to have been set to unity . ] : @xmath7 the formalism has been extended by allowing the cosmological constant of the theory to be dynamical , supplying a pressure _ via _
@xmath8 , along with its conjugate volume @xmath9 .
now , the black hole mass is related to the enthalpy @xmath10 of the system instead of the energy @xmath4@xcite : @xmath11 .
the first law now reads as : @xmath12 the temperature and the entropy remain related to the surface gravity and area of the black hole as in equations ( [ oldtd ] ) .
the @xmath13 are gauge charges , and @xmath14 are angular momenta , while @xmath15 and @xmath16 are their conjugate potentials and angular velocities , respectively .
the black holes may have other parameters and they enter additively with their conjugates to the first law ( [ newtd ] ) in the usual way .
this formalism works in multiple dimensions .
interestingly , for the static black holes , the thermodynamic volume @xmath9 is just the naive `` geometric '' volume of the black holes : the volume of the ball of radius @xmath17 ( our notation for the horizon radius in this paper ) . in this
_ extended _ black hole thermodynamics , since the pressure and volume are now in play , alongside temperature and entropy ( [ newtd ] ) , it is natural to study devices which can extract useful mechanical work from heat energy , _
i.e. _ , traditional heat engines@xcite .
these devices were named `` holographic heat engines '' , since for negative cosmological constant ( _ i.e. _ with positive pressure , since @xmath18 ) such cycles represent a journey through a family of holographically dual@xcite non gravitational field theories ( at large @xmath19 ) defined in one dimension fewer .
although we have holographic applications in mind for some of this work , for this paper our focus will be on the black hole side of the story , an interesting context in its own right .
so for the purposes of the gravitational theory , the working substance of the heat engine is a particular black hole solution of the gravity system .
it supplies an equation of state through the relation between its temperature @xmath5 and the black hole parameters defined in the usual way ( we will give examples below ) .
the precise form of all these relations depends on the type of black hole , and the parent theory of gravity under discussion .
one may extract mechanical work from such an engine@xcite _ via _ the @xmath20 term in the first law of thermodynamics in the classic way : define a closed cycle in state space during which there is a net input heat flow @xmath21 , a l0.4 net output heat flow @xmath22 , and a net output work @xmath23 .
so @xmath24 .
a central quantity , the efficiency for the cycle , is defined as @xmath25 .
its value is sensitive to the details of the equation of state of the system and also to the choice of cycle in state space .
consider the cycle given in figure [ fig : prototype ] . in refs.@xcite
it is explained why this is a natural choice for static black holes . for such holes ,
the entropy and the volume are not independent , being both simple functions of @xmath26 , the horizon radius .
so isochores are adiabats , and so the only heat flows are along the top and bottom lines .
computing the efficiency boils down to evaluating @xmath27 along those isobars , where @xmath28 is the specific heat at constant pressure . in general , calculation of efficiency
is a difficult task to perform exactly using this approach , and high temperature or high pressure computations are used to get approximate results@xcite .
recently , however , ref .
@xcite showed a much simpler way to evaluate the efficiency .
the first law is : @xmath29 and along the isobars , @xmath30 .
therefore the total heat flow along an isobar is simply the enthalpy change .
normally , that might not be a useful rewriting , but in extended gravitational thermodynamics , a precise expression for the enthalpy is readily available since it is just the black hole mass @xmath1 .
this results in a remarkably simple exact formula : @xmath31 where the black hole mass is evaluated at each corner of the rectangle , with the labelling given in figure [ fig : prototype ] .
@xmath1 is usually written as a function of @xmath17 and @xmath32 . in the examples of this paper ,
since @xmath9 is a simple function of @xmath17 , we will be easily able to write down @xmath1 as a function of @xmath32 and @xmath9 .
it was also shown in ref.@xcite that the result ( [ simplemass ] ) can be used as the basis for an algorithm for computing the efficiency of a cycle of _ arbitrary _ shape to any desired accuracy . any closed shape on the state space
can be approximated by tiling with a regular lattice of rectangles .
r0.4 this is possible because cycles are additive ( see figure [ fig : addlaw ] ) .
consequently , only the cells at the edge contribute .
any mismatch between the edge of the cycle s contour and the tiling s edge can be reduced by simply shrinking the size of the unit cell .
edge cells are called hot cells if they have their upper edges open , and cold cells if they have their lower edges open .
summing all the hot cell mass differences ( evaluated at the top edges ) will give @xmath21 and summing all the cold cell mass differences ( evaluated at the bottom edges ) will yield @xmath22 . so the efficiency is : @xmath33 where we have labelled all cells corners in the same way as the prototype cycle in figure [ fig : prototype ] .
an example with a triangular cycle was given in @xcite to show the algorithm in action supporting the previous argument .
as already stated , a given black hole , thought of as a working substance for a heat engine , supplies a particular equation of state .
the efficiency will depend upon this choice .
moreover , the efficiency will also depend upon the details of the choice of cycle . for maximizing @xmath34 ,
certain choices of cycle will be better adapted to a particular working substance ( choice of black hole ) than others .
( for example , for the same cycle of figure [ fig : prototype ] , a non - static black hole will generically have a larger @xmath35 due to non zero heat flows along the isochores , and therefore a smaller @xmath34 . )
so a natural question arises : how does one compare the efficiency of different working substances ?
we have in mind a comparison that depends as little as possible on special choices of cycle .
in other words , in comparing working substances for making a heat engine , we should _ not _ choose a special cycle that favours one black hole s particular properties over another .
notice that this requirement requires us to make a choice that is in opposition to what is normally done : cycles are usually chosen in a way that is naturally adapted to the equation of state in order to simplify computation .
so we are asking that a more difficult choice of cycle be made , by necessity .
this is where the exact formula and algorithm reviewed above come in .
we can pick a benchmark cycle of whatever shape seems appropriate and implement the algorithm to compute @xmath34 to any desired accuracy .
this freedom allows us to make the following choice of benchmark : we choose the cycle to be a _ circle _ in the @xmath0 plane .
the logic of this choice is that the circle is a simply parametrised shape which is also unlikely to favour _ any _ species of black hole ( working substance ) whatsoever . no thermodynamic variable is unchanged on any segment of the cycle , so it is , in some sense , a difficult cycle for all black holes .
all that needs be specified is the origin of the circle and its radius .
these properties make it an excellent choice of benchmark .
the outline of this paper is as follows . in section [ sec : circle ]
, we set up the circular cycle as our benchmarking tool , and explain our implementation of the exact formula and algorithm of ref.@xcite for calculating the efficiency .
we then discuss , in section [ sec : ideal ] , a very special case of working substance : an ideal gas " like system .
it allows us to derive some exact results that help test our implementation , and which also set a new benchmark standard for later use . in section [ sec : comparison ] , we compare three examples of black holes as working substances for heat engines : charged ( reissner nordstrom like ) black holes , gauss bonnet black holes , and born infeld black holes .
we conclude in section [ sec : conclusions ] with a brief discussion of future applications of our benchmarking procedure .
for our circle , we implemented the algorithm and exact formula of ref.@xcite , with the aid of a computer , as follows : imagine that we have chosen the origin and radius , @xmath36 , of the circle in the @xmath0 plane .
we next overlaid it onto the @xmath37 regular lattice of squares of total side length @xmath38 . for simplicity , we used even @xmath39 so that there are same number of squares both in the upper half and in the lower half of the circle . next we computed the pressure and volume at each corner of all the squares . using simple geometry ,
we determined which squares intersect the circle .
we checked for cases where two squares share a common isobar and both intersect the circle .
then , if we are in the upper part of the circle , we remove the one below and keep only the upper square .
we did this check in the lower half of the circle in a similar fashion .
this allowed us to identify all the hot cells and cold cells of the approximation , and their @xmath40 coordinates .
the black hole mass is a function of pressure @xmath32 and volume @xmath9 only ( with some parameters that we have already fixed ) , so we can compute its value at each corner .
then we use the formula ( [ fullmass ] ) to give us the approximate @xmath34 for that level of granularity . increasing the value of @xmath39
makes the size of the unit cell smaller , making the path traced by the hot and cold cells a better fit to our circle , reducing the error in @xmath34 .
indeed , we found that just as for the triangle prototype of ref.@xcite , the efficiency converges nicely for large @xmath39 .
( see the examples in section [ sec : comparison ] . )
( 100 squares ) .
red lines are the tops of hot cells and blue lines are the bottoms of cold cells .
as @xmath39 increases , these lines converge to the boundary of the circle .
the dashed black lines are sample isotherms.,width=336 ] we can even do more . since temperature is also a function of @xmath32 and @xmath9 , we can compute it at each corner . then while we run over all the cells to compute @xmath34 , we can keep track of the maximum and the minimum temperatures ( @xmath41 and @xmath42 ) achieved in the entire cycle .
hence , we can compute the carnot efficiency @xmath43 for this engine .
this will be a check of our results because no cycle can have a greater efficiency than a carnot cycle .
figure [ fig : ccn10 ] shows an example for @xmath44 .
the green crosses show the points of the square lattice .
the circle is our circular cycle .
red segments are the tops of the hot cells and blue segments are the bottoms of cold cells .
the black dashed lines show a few sample isotherms determined from the underlying equation of state of the system in question .
( this example is a snapshot of the einstein hilbert
maxwell case more fully explored in section [ ehm ] ) . in choosing our benchmark cycle to compare different black holes
, we should fix the circle origin and radius @xmath36 .
generically , the choices do nt matter , as long as they are the same across the comparison .
we chose @xmath45 here , and in the following sections , purely arbitrarily , except for making sure that we avoided any regions where the equations of state of the black holes under comparison had any multi valuedness that would signal non trivial phase transitions@xcite .
such regimes require a separate , more careful study in this heat engine context that are beyond the scope of this paper .
one might worry that since the circular cycle is presumably not even close to a cycle for which one has an analytically computable result , if there was an error , it might not be noticed .
the carnot test above is useful , but it is a rather weak upper bound on the efficiency .
we derive some complementary tests , and a stronger ( exact ) bound , in the next section .
before we proceed to study some black hole examples we briefly pause to study a simple but instructive case .
it is in fact a limiting form of all of the black hole solutions we ll discuss shortly . as discussed in ref.@xcite it deserves to be called an ideal gas case , and as such , sets an additional standard by which we might assess other working substances . in dimension @xmath46 ,
the leading large horizon radius ( @xmath26 ) limit of all the asymptotically anti
de sitter black holes we will discuss is rather simple , with dependence for the mass and temperature as follows : @xmath47 where @xmath48 is the volume ( _ i.e. , _ surface area ) of the unit round @xmath49 sphere .
the exact thermodynamic volume for all of the static black holes under study is : @xmath50 and so we have the familiar ideal gas " behaviour in this large @xmath26 limit : @xmath51 a family of hyperbolae in the @xmath52 plane where @xmath53 . this ideal gas can be obtained as a limit for any of our black holes ( in later sections ) as either a large @xmath26 limit or as a high temperature limit . before moving on to those cases
, we can study this in its own right , taking the above as the equation of state everywhere in the @xmath0 plane .
notice first that the efficiency of any cell such as the prototype of figure [ fig : prototype ] simplifies nicely in this case .
this is because the mass is simply @xmath54 , and hence @xmath32 factors out in each mass difference , leaving only a volume difference .
so @xmath34 for figure [ fig : prototype ] is just@xcite @xmath55 . turning to the efficiency of the circle
, the factorization into sums of volume differences means that there is no dependence of the result on the volume coordinate of the circle s origin : any shift in the origin will cancel out everywhere .
we can say even more in this case however .
in fact , the terms in the sums in the algorithm ( [ fullmass ] ) are actually entirely geometrical in interpretation ! for example , for a hot cell a term is of the form @xmath56 .
this is simply the area of the rectangular strip that starts on the @xmath9 axis and is bounded above by the top of the cell .
this is a clue to writing an _ exact _
formula for the efficiency in the case of our ideal gas .
the simplest way to do it is to rewrite @xmath34 as the ratio of work to heat flowing in , @xmath57 .
now @xmath58 while from our observation above , @xmath35 is , in the large @xmath39 limit , exactly the area underneath the upper semi circle of the circular path : @xmath59 , so our result is : @xmath60 where @xmath32 is the pressure at the centre of the circle .
this exact formula is rather surprising .
notably , in addition to being independent of @xmath9 it is also independent of spacetime dimension , but the real surprise is that the algorithm assembled itself into a purely geometric result that yielded an exact formula for what is , on the face of it , a difficult shape of cycle .
in fact , this exact geometrical result will work for _ any cycle shape_. perhaps there can be other surprises of this sort for other systems besides this special ideal gas case .
the formula is also a rather useful check on our methods for a number of reasons .
the first is that the @xmath32 and @xmath36 dependence are non trivial predictions , and so we were obliged to check to see if our discrete algorithm reproduces such dependence , and indeed it did .
for example , figure [ fig : effrad ] shows , for @xmath61 , some example points computed by inserting the ideal gas into our algorithm .
the red curve is the exact result of equation ( [ eq : idealgasefficiency ] ) . .
the ideal gas ( see text ) equation of state was used in the algorithm for @xmath61 with the circle origin at @xmath62 , and radius @xmath63 .
the blue crosses plot the result for @xmath34 .
the red curve is a plot of the exact result from equation ( [ eq : idealgasefficiency]).,width=240 ] the second reason this is a strong check is that it presents a lower upper bound on our results than the upper bound given by carnot ( discussed in the previous section ) . our black holes , in the regions where we study them ,
can be thought of as perturbations of this ideal gas case , and so we should expect that the efficiencies we obtain approach ( but do not exceed ) the ideal gas result .
we have , for the comparisons to come , the circle s origin at @xmath64 , @xmath65 , and its radius as @xmath63 , for which the ideal gas efficiency is ( to six significant figures ) @xmath66 .
it is worth noting that using the discretisation algorithm to compute the ideal gas case gives @xmath67 at @xmath61 and @xmath68 at @xmath69 .
( moving significantly beyond @xmath69 to see further convergence proved beyond the numerical capabilities of the system we were using . )
we now apply our benchmark cycle to a sampling of different black holes acting as working substances .
we will only briefly introduce the black holes since they are well known in the literature .
they were used in heat engines in refs.@xcite , with some analysis and comparison presented there , but now we have a clearer , more systematic benchmarking procedure .
we will work in @xmath70 for definiteness ( it is trivial to insert the formulae for other dimensions into our algorithm ; we saw no compelling reason to present the results for other dimensions here ) , and our benchmark circle will be centred at @xmath64 , @xmath65 , with radius @xmath63 . in each case
we list the bulk action in @xmath70 dimensions and the mass and temperature of the black hole . for static black holes
the volume @xmath9 is simply : @xmath71 . also , recall that the cosmological constant @xmath72 is related to pressure @xmath32 _ via _
@xmath18 , and in @xmath46 dimensions @xmath72 sets a length scale @xmath73 through @xmath74 .
so in @xmath70 dimensions , @xmath75 .
the mass and temperature formulae we present will have had @xmath73 eliminated in favour of @xmath32 .
note that in presenting our results for the efficiency , the engine s actual efficiency will be denoted by @xmath34 ( without a subscript ; the surrounding text will make it clear which case is being discussed ) and the associated carnot efficiency will be denoted @xmath76 ( again with context making it clear as to which case is being discussed ) .
this will help us avoid a proliferation of subscripts . the bulk action for the einstein hilbert
maxwell system in @xmath70 is : @xmath77 . here
maxwell black holes are used as the working substance .
blue crosses represent the carnot efficiency @xmath76 , while black squares represent @xmath34 . for @xmath61 , @xmath76 and @xmath34
converge to 0.6674942748 and 0.5653677678 respectively.,width=288 ] we can now write the mass and the temperature of the einstein hilbert maxwell ( _ i.e. , _ reissner nordstrom like ) black hole solution , parametrized by a charge @xmath78 ( which we will later choose as @xmath79 ) : @xmath80 and we can write them entirely in terms of @xmath32 and @xmath9 , using @xmath81 . figure [ fig : emn500 ] shows the results of the algorithm for computing @xmath76 and @xmath34 for the benchmark circle in this case . in the presence of a gauss bonnet sector ,
the action becomes : @xmath82 where @xmath83 is the gauss bonnet parameter which has dimensions of @xmath84 .
if we set @xmath85 in ( [ gbaction ] ) we go back to the previous case of einstein hilbert maxwell system ( [ ehmaction ] ) . . here
bonnet black holes are used as the working substance .
blue crosses represent the carnot efficiency @xmath76 , while black squares represent @xmath34 . for @xmath61 , @xmath76 and @xmath34
converge to 0.6674954523 and 0.5653678245 respectively.,width=288 ] the mass and temperature of the black hole , parametrized by @xmath78 and @xmath86 are : @xmath87 where @xmath88 .
we will again work with @xmath89 and we choose a sample value of the coupling as @xmath90 .
see figure [ fig : gbn500 ] for @xmath76 and @xmath34 from the benchmark analysis .
the so called born infeld action @xcite is a non - linear generalization of the maxwell action , controlled by the parameter @xmath91 : @xmath92 if we take the limit @xmath93 in ( [ bisector ] ) we recover old maxwell action . the einstein hilbert born infeld bulk action in @xmath70 is obtained by replacing the maxwell sector in equation [ ehmaction ] with this action . .
here born
infeld black holes are used as the working substance .
blue crosses represent the carnot efficiency @xmath76 , while black squares represent @xmath34 . for @xmath61 , @xmath76 and @xmath34
converge to 0.6674942730 and 0.5653678967 respectively.,width=288 ] the exact results for the born infeld black hole s mass and temperature are known , but for our purposes , it is enough to expand them in @xmath94 , keeping only leading non trivial terms . for the mass : @xmath95 and the temperature : @xmath96 the exact formulae are computationally intensive , and for any significant @xmath39 , there are far too many computations to allow computation of the efficiency in a reasonable amount of time ( especially in @xmath70 ) and so we chose to make this truncation at the outset .
we worked with @xmath89 and @xmath97 in our benchmark studies , the results of which are shown in figure [ fig : bin500 ] for @xmath76 and @xmath34 . in figure
[ fig : pc ] we gather all the efficiencies computed using the benchmark cycle together .
the gauss bonnet and born infeld cases , thought of as perturbations of the einstein
maxwell case , have higher efficiencies , although it is interesting that the differences begin to show only in the 8th significant figure , for the parameter values chosen for @xmath86 and @xmath91 .
we explored other parameter values ( while making sure to stay in the physical range allowed by reality of the mass for the gauss bonnet case ) and found a very weak dependence of the efficiency as they varied .
( this all matches observations made in refs.@xcite in the high temperature limit . )
they all in turn have significantly lower efficiency than the ideal gas case listed at the end of section [ sec : ideal ] . with circle origin at @xmath62 and radius @xmath63 .
for additional comparison , the ideal gas case of section [ sec : ideal ] has @xmath98 at @xmath61 , and is @xmath99 exactly.,width=192 ]
we ve defined a new way of comparing different black holes , given meaning in the context of defining black hole heat engines@xcite in extended thermodynamics .
our benchmarking allowed us to compare four important cases against each other , and we found results consistent with earlier studies reported in refs.@xcite , but here we ve established a more robust framework for comparison ( a standard circular cycle ) facilitated by the exact formula and algorithm of ref.@xcite . along the way , we found a fascinating case where the algorithm itself collapses to another exact result , this time the exact efficiency of an
ideal gas " example
. it would be fascinating to see if other exact results of this kind can be obtained for other non trivial systems
. it would be interesting to study other black holes using this same benchmarking scheme in order to compare more properties of heat engine working substances .
extend all this to non
static cases would be particularly worthwhile .
finally , the possible applications of all of this to holographically dual strongly coupled field theories is worth exploring .
we hope to report on some of this elsewhere .
we would like to thank the us department of energy for support under grant de
cvj would like to thank amelia for her patience and support .
j. d. bekenstein , `` black holes and entropy , '' http://dx.doi.org/10.1103/physrevd.7.2333[_phys.rev._ * d7 * ( 1973 ) 23332346 ] . j. d. bekenstein , `` generalized second law of thermodynamics in black hole physics , '' http://dx.doi.org/10.1103/physrevd.9.3292[_phys.rev._ * d9 * ( 1974 ) 32923300 ] .
s. hawking , `` particle creation by black holes , '' http://dx.doi.org/10.1007/bf02345020[_commun.math.phys._ * 43 * ( 1975 ) 199220 ] . s. hawking , `` black holes and thermodynamics , '' http://dx.doi.org/10.1103/physrevd.13.191[_phys.rev._ * d13 * ( 1976 ) 191197 ] .
m. m. caldarelli , g. cognola , and d. klemm , `` thermodynamics of kerr - newman - ads black holes and conformal field theories , '' http://dx.doi.org/10.1088/0264-9381/17/2/310[_class.quant.grav._ * 17 * ( 2000 ) 399420 ] , http://arxiv.org/abs/hep-th/9908022[arxiv:hep-th/9908022 [ hep - th ] ] .
s. wang , s .- q .
wu , f. xie , and l. dan , `` the first laws of thermodynamics of the ( 2 + 1)-dimensional btz black holes and kerr - de sitter spacetimes , '' http://dx.doi.org/10.1088/0256-307x/23/5/009[_chin.phys.lett._ * 23 * ( 2006 ) 10961098 ] , http://arxiv.org/abs/hep-th/0601147[arxiv:hep-th/0601147 [ hep - th ] ] .
y. sekiwa , `` thermodynamics of de sitter black holes : thermal cosmological constant , '' http://dx.doi.org/10.1103/physrevd.73.084009 [ _ phys.rev . _ * d73 * ( 2006 ) 084009 ] , http://arxiv.org/abs/hep-th/0602269[arxiv:hep-th/0602269 [ hep - th ] ] . e. a. larranaga rubio , `` stringy generalization of the first law of thermodynamics for rotating btz black hole with a cosmological constant as state parameter , '' http://arxiv.org/abs/0711.0012[arxiv:0711.0012 [ gr - qc ] ] .
d. kastor , s. ray , and j. traschen , `` enthalpy and the mechanics of ads black holes , '' http://dx.doi.org/10.1088/0264-9381/26/19/195011 [ _ class.quant.grav . _
* 26 * ( 2009 ) 195011 ] , http://arxiv.org/abs/0904.2765[arxiv:0904.2765 [ hep - th ] ] .
b. p. dolan , `` the cosmological constant and the black hole equation of state , '' http://dx.doi.org/10.1088/0264-9381/28/12/125020 [ _ class.quant.grav .
_ * 28 * ( 2011 ) 125020 ] , http://arxiv.org/abs/1008.5023[arxiv:1008.5023 [ gr - qc ] ] .
m. cvetic , g. gibbons , d. kubiznak , and c. pope , `` black hole enthalpy and an entropy inequality for the thermodynamic volume , '' http://dx.doi.org/10.1103/physrevd.84.024037[_phys.rev._ * d84 * ( 2011 ) 024037 ] , http://arxiv.org/abs/1012.2888[arxiv:1012.2888 [ hep - th ] ] .
b. p. dolan , `` compressibility of rotating black holes , '' http://dx.doi.org/10.1103/physrevd.84.127503[_phys.rev._ * d84 * ( 2011 ) 127503 ] , http://arxiv.org/abs/1109.0198[arxiv:1109.0198 [ gr - qc ] ] . b. p. dolan , `` pressure and volume in the first law of black hole thermodynamics , '' http://dx.doi.org/10.1088/0264-9381/28/23/235017 [ _ class.quant.grav .
_ * 28 * ( 2011 ) 235017 ] , http://arxiv.org/abs/1106.6260[arxiv:1106.6260 [ gr - qc ] ] .
b. p. dolan , `` where is the pdv term in the fist law of black hole thermodynamics ?
, '' http://arxiv.org/abs/1209.1272[arxiv:1209.1272 [ gr - qc ] ]
. n. altamirano , d. kubiznak , r. b. mann , and z. sherkatghanad , `` thermodynamics of rotating black holes and black rings : phase transitions and thermodynamic volume , '' http://dx.doi.org/10.3390/galaxies2010089[_galaxies_ * 2 * ( 2014 ) 89159 ] , http://arxiv.org/abs/1401.2586[arxiv:1401.2586 [ hep - th ] ] .
d. kubiznak , r. b. mann , and m. teo , `` black hole chemistry : thermodynamics with lambda , '' http://arxiv.org/abs/1608.06147[arxiv:1608.06147 [ hep - th ] ] .
m. henneaux and c. teitelboim , `` the cosmological constant as a canonical variable , '' http://dx.doi.org/10.1016/0370-2693(84)91493-x[_phys.lett._ * b143 * ( 1984 ) 415420 ] . c. teitelboim ,
`` the cosmological constant as a thermodynamic black hole parameter , '' http://dx.doi.org/10.1016/0370-2693(85)91186-4[_phys.lett._ * b158 * ( 1985 ) 293297 ] .
m. henneaux and c. teitelboim , `` the cosmological constant and general covariance , '' http://dx.doi.org/10.1016/0370-2693(89)91251-3[_phys.lett._ * b222 * ( 1989 ) 195199 ] .
m. k. parikh , `` the volume of black holes , '' http://dx.doi.org/10.1103/physrevd.73.124021[_phys.rev._ * d73 * ( 2006 ) 124021 ] , http://arxiv.org/abs/hep-th/0508108[arxiv:hep-th/0508108 [ hep - th ] ] . c. v. johnson , `` holographic heat engines , '' http://dx.doi.org/10.1088/0264-9381/31/20/205002[_class .
* 31 * ( 2014 ) 205002 ] , http://arxiv.org/abs/1404.5982[arxiv:1404.5982 [ hep - th ] ] .
j. m. maldacena , `` the large n limit of superconformal field theories and supergravity , '' _ adv .
* 2 * ( 1998 ) 231252 , http://arxiv.org/abs/hep-th/9711200[hep-th/9711200 ] .
e. witten , `` anti - de sitter space and holography , '' _ adv .
* 2 * ( 1998 ) 253291 , http://arxiv.org/abs/hep-th/9802150[hep-th/9802150 ] .
s. s. gubser , i. r. klebanov , and a. m. polyakov , `` gauge theory correlators from non - critical string theory , '' _ phys .
_ * b428 * ( 1998 ) 105114 , http://arxiv.org/abs/hep-th/9802109[hep-th/9802109 ] .
e. witten , `` anti - de sitter space , thermal phase transition , and confinement in gauge theories , '' _ adv .
_ * 2 * ( 1998 ) 505532 , http://arxiv.org/abs/hep-th/9803131[hep-th/9803131 ] .
o. aharony , s. s. gubser , j. m. maldacena , h. ooguri , and y. oz , `` large n field theories , string theory and gravity , '' http://dx.doi.org/10.1016/s0370-1573(99)00083-6[_phys . rept . _
* 323 * ( 2000 ) 183386 ] , http://arxiv.org/abs/hep-th/9905111[arxiv:hep-th/9905111 [ hep - th ] ] . c. v. johnson , `` gauss - bonnet black holes and holographic heat engines beyond large n , '' http://arxiv.org/abs/1511.08782[arxiv:1511.08782 [ hep - th ] ] . c. v. johnson , `` born - infeld ads black holes as heat engines , '' http://arxiv.org/abs/1512.01746[arxiv:1512.01746 [ hep - th ] ] .
a. belhaj , m. chabab , h. el moumni , k. masmar , m. b. sedra , and a. segui , `` on heat properties of ads black holes in higher dimensions , '' http://dx.doi.org/10.1007/jhep05(2015)149[_jhep_ * 05 * ( 2015 ) 149 ] , http://arxiv.org/abs/1503.07308[arxiv:1503.07308 [ hep - th ] ] . j. sadeghi and k. jafarzade , `` heat engine of black holes , '' http://arxiv.org/abs/1504.07744[arxiv:1504.07744 [ hep - th ] ] .
e. caceres , p. h. nguyen , and j. f. pedraza , `` holographic entanglement entropy and the extended phase structure of stu black holes , '' http://dx.doi.org/10.1007/jhep09(2015)184[_jhep_ * 09 * ( 2015 ) 184 ] , http://arxiv.org/abs/1507.06069[arxiv:1507.06069 [ hep - th ] ] .
m. r. setare and h. adami , `` polytropic black hole as a heat engine , '' http://dx.doi.org/10.1007/s10714-015-1979-0[_gen .
_ * 47 * ( 2015 ) no . 11 , 133 ] . m. zhang and w .- b . liu , `` f(r ) black holes as heat engines , '' http://dx.doi.org/10.1007/s10773-016-3134-4[_int . j. theor . phys . _ * 55 * ( 2016 ) no . 12 , 51365145 ] . c. bhamidipati and p. k. yerra , `` heat engines for dilatonic born - infeld black holes , '' http://arxiv.org/abs/1606.03223[arxiv:1606.03223 [ hep - th ] ] .
wei and y .- x .
liu , `` implementing black hole as efficient power plant , '' http://arxiv.org/abs/1605.04629[arxiv:1605.04629 [ gr - qc ] ] .
j. sadeghi and k. jafarzade , `` the modified horava - lifshitz black hole from holographic engine , '' http://arxiv.org/abs/1604.02973[arxiv:1604.02973 [ hep - th ] ] . c. v. johnson , `` an exact efficiency formula for holographic heat engines , '' http://dx.doi.org/10.3390/e18040120[_entropy_ * 18 * ( 2016 ) 120 ] , http://arxiv.org/abs/1602.02838[arxiv:1602.02838 [ hep - th ] ] .
a. chamblin , r. emparan , c. v. johnson , and r. c. myers , `` charged ads black holes and catastrophic holography , '' _ phys .
_ * d60 * ( 1999 ) 064018 , http://arxiv.org/abs/hep-th/9902170[hep-th/9902170 ] .
a. chamblin , r. emparan , c. v. johnson , and r. c. myers , `` holography , thermodynamics and fluctuations of charged ads black holes , '' _ phys .
_ * d60 * ( 1999 ) 104026 , http://arxiv.org/abs/hep-th/9904197[hep-th/9904197 ] .
d. kubiznak and r. b. mann , `` p - v criticality of charged ads black holes , '' http://dx.doi.org/10.1007/jhep07(2012)033[_jhep_ * 1207 * ( 2012 ) 033 ] , http://arxiv.org/abs/1205.0559[arxiv:1205.0559 [ hep - th ] ] .
cai , l .- m .
cao , l. li , and r .- q .
yang , `` p - v criticality in the extended phase space of gauss - bonnet black holes in ads space , '' http://dx.doi.org/10.1007/jhep09(2013)005[_jhep_ * 09 * ( 2013 ) 005 ] , http://arxiv.org/abs/1306.6233[arxiv:1306.6233 [ gr - qc ] ] .
m. born , `` modified field equations with a finite radius of the electron , '' http://dx.doi.org/10.1038/132282a0[_nature_ * 132 * ( 1933 ) 282282 ] .
m. born , `` quantum theory of the electromagnetic field , '' http://dx.doi.org/10.1098/rspa.1934.0010[_proc .
lond . _ * a143 * ( 1934 ) 410437 ] .
m. born and l. infeld , `` foundations of the new field theory , '' http://dx.doi.org/10.1098/rspa.1934.0059[_proc .
lond . _ * a144 * ( 1934 ) 425451 ] .
s. fernando and d. krug , `` charged black hole solutions in einstein - born - infeld gravity with a cosmological constant , '' http://dx.doi.org/10.1023/a:1021315214180[_gen .
* 35 * ( 2003 ) 129137 ] , http://arxiv.org/abs/hep-th/0306120[arxiv:hep-th/0306120 [ hep - th ] ] .
cai , d .- w .
pang , and a. wang , `` born - infeld black holes in ( a)ds spaces , '' http://dx.doi.org/10.1103/physrevd.70.124034[_phys .
_ * d70 * ( 2004 ) 124034 ] , http://arxiv.org/abs/hep-th/0410158[arxiv:hep-th/0410158 [ hep - th ] ] .
t. k. dey , `` born - infeld black holes in the presence of a cosmological constant , '' http://dx.doi.org/10.1016/j.physletb.2004.06.047 [ _ phys .
lett . _ * b595 * ( 2004 ) 484490 ] , http://arxiv.org/abs/hep-th/0406169[arxiv:hep-th/0406169 [ hep - th ] ] . | we present the results of initiating a benchmarking scheme that allows for cross comparison of the efficiencies of black holes used as working substances in heat engines .
we use a circular cycle in the @xmath0 plane as the benchmark engine .
we test it on einstein maxwell , gauss bonnet , and born infeld black holes . also , we derive a new and surprising exact result for the efficiency of a special ideal gas " system to which all the black holes asymptote .
arxiv:1612 * avik chakraborty and clifford v. johnson * _ department of physics and astronomy _ _ university of southern
california _
_ los angeles , ca 90089 - 0484 , u.s.a .
_ avikchak , johnson1 , [ at ] usc.edu |
one of the open questions in neutrino physics is the mass ordering of the neutrinos ; whether they are ordered normally or inverted .
many experiments intend to determine the mass ordering , of which the 50 kton iron calorimeter ( ical ) detector at the proposed india - based neutrino observatory is one ambitious experiment @xcite .
ical will be a magnetised iron calorimeter mainly sensitive to muons produced in the charged - current ( cc ) interactions of atmospheric muon neutrinos ( and anti - neutrinos ) with the iron target in the detector .
it can distinguish cc muon - neutrino - induced events from anti - neutrino - induced ones since the former interaction produces @xmath10 while the latter produces @xmath11 in the detector and ical has excellent muon charge identification ( cid ) capability .
this is also crucial to determine precisely the momentum of the muons through bending in the magnetic field .
since matter effects are different between neutrino and anti - neutrino propagation in the earth , this feature can help resolve the neutrino mass ordering by determining the sign of the 23 mass - squared difference @xmath12 , @xmath13 , being the neutrino mass eigenstates @xcite .
in addition , the matter effects improve the sensitivity to the magnitude @xmath14 of this mass - squared difference as well as to the 23 mixing angle , @xmath15 , provided the across - generation mixing angle @xmath16 is rather well - known , which is indeed the case @xcite .
many previous analyses have been reported , projecting the sensitivities of ical detector to oscillation parameters in the 23 sector @xcite as also the mass ordering .
the sensitivity to mass hierarchy is directly proportional to the value of @xmath16 which is quite precisely known @xcite .
it also depends on the ability of ical to separate neutrino and anti - neutrino events which is possible since ical is magnetised . while there is an uncertainty of about 20% on the atmospheric neutrino fluxes themselves , the uncertainty on their _ ratios
_ is much smaller , about 5% , and was ignored in earlier analyses @xcite . in this paper , we show that this smaller uncertainty on the ratio acts as a constraint that in turn significantly shrinks the allowed parameter space , especially for @xmath3 . for instance , we will see that the precision on @xmath3 decreases from 13% to 9% in a certain analysis mode when this constraint is included .
this is generally true for all magnetised detectors . to our knowledge
such an effect has not been discussed in the literature earlier .
the paper is structured as follows .
all the results in this paper are with detailed simulation studies of the physics processes at the ical detector .
the main steps involved in this are neutrino event generation , inclusion of detector responses and efficiencies , inclusion of oscillations , binning in observables and @xmath2 analysis .
the procedure of neutrino event generation with the nuance neutrino @xcite generator and the implementation of oscillations are discussed in detail in section [ nu - evt ] .
the choice of observables and kinematic regions used in the analysis , along with the inclusion of detector responses are discussed in detail in section [ obs ] .
the effect of increasing the energy range of observed muons is also explained in this section .
the detailed @xmath2 analysis and a discussion of the systematic errors that have been considered are presented in section [ chisq ] .
the results of precision measurements and hierarchy sensitivity studies are shown in section [ results ] .
the impact of the additional pull in the @xmath17 flux ratio implemented in this analysis is discussed in detail in section [ eff-11-pull ] . the summary and conclusions are given in section [ summary ] .
the interactions of interest in ical are the cc interactions of @xmath0 and @xmath1 with the iron target in ical .
these @xmath0 ( @xmath1 ) in ical come from both @xmath0 and @xmath18 atmospheric fluxes via @xmath19 and @xmath20 oscillations .
the first channel gives the number of @xmath0 events which have survived and the second , subdominant , one gives the number from oscillations of @xmath18 to @xmath0 .
the number of events ical sees will be a sum of these events . thus , @xmath21 \times \frac{d\sigma_\mu ( e_\nu)}{d e_\mu d\cos\theta_\mu}~ , \label{toteve}\end{aligned}\ ] ] where @xmath22 is the number of target nucleons in the detector , @xmath23 is the differential neutrino interaction cross section in terms of the energy and direction of the cc lepton produced , @xmath24 and @xmath25 are the @xmath0 and @xmath18 fluxes and @xmath26 is the oscillation probability of @xmath27 . the number of _ unoscillated _ events over an exposure time @xmath28 in a bin of ( @xmath29 ) is obtained from the nuance neutrino generator using the honda 3d atmospheric neutrino fluxes @xcite , neutrino - nucleus cross - sections , and a simplified ical detector geometry . while nuance lists details of all the final state particles including the muon and all hadrons , ical will be optimised to determine accurately the energy and direction of the muons ( seen as a clean track in the detector ) and the summed energy of all the hadrons in the final state ( since it can not distinguish individual hadrons ) . even though the analyses are done for a smaller number of years ( say 10 ) , a huge sample of nuance events for a very large number of years ( here 1000 years ) is generated and scaled down to the required number of years during the analysis .
this is mainly done to reduce the effect of statistical ( monte carlo ) fluctuations on sensitivity studies , which may alter the results .
a detailed discussion about the effect of fluctuations on oscillation sensitivity studies will be discussed in appendix [ fluct ] .
a sample of 1000 years of unoscillated events was generated using nuance .
two sets were generated : 1 .
cc muon events using the @xmath24 flux and 2 .
cc muon events obtained by swapping @xmath30 fluxes .
this generates the so - called muon- and swapped - muon events that correspond to the two terms in eq .
[ toteve ] .
these events are oscillated depending on the neutrino oscillation parameters being used .
the oscillation probabilities are calculated by considering the full three flavour oscillations in the presence of matter effects .
the preliminary reference earth model ( prem ) profile @xcite has been used to model the varying earth matter densities encountered by the neutrinos during their travel through the earth .
the runge - kutta solver method is used to calculate the oscillation probabilities @xcite for various energies @xmath31 and distances @xmath32 , or equivalently , @xmath33 ( @xmath34 being the zenith angle ) of the neutrino .
further discussion of the oscillation probabilities and plots of a few sample curves are presented in the next section after listing the kinematical range of interest .
the oscillation is applied event by event ( for both muon and swapped muon events ) as discussed in detail in ref .
@xcite and it is a time consuming process since the actual sample contains 1000 years of events .
the central values of the oscillations parameters are given in table [ osc - par-3sig ] along with their known @xmath35 range .
note that @xmath36 is currently unknown and its true value has been assumed to be @xmath37 for the purposes of this calculation .
furthermore , since ical is insensitive @xcite to this parameter , it has been kept fixed in the calculation , along with the values of the 12 oscillation parameters @xmath38 and @xmath39 which also do not affect the results .
.main oscillation parameters used in the current analysis . in the second column are the true values of these parameters used to simulate the `` observed '' data set .
true value is the value at which the data is simulated .
more details are given in the main text .
for precision measurement of each parameter , all others are varied except that parameter in the analysis . [ cols="^,^,^",options="header " , ] the results discussed in the previous sections were for fixed values of either of the oscillation parameters @xmath3 and @xmath42 .
we now discuss the parameter space allowed by our latest analysis .
the analysis was done for the 11 pull case with @xmath43 = 0.525 gev and with hadrons , for 500 kton yrs of ical exposure .
normal hierarchy is assumed to be the true hierarchy .
the 90% and 99% confidence contours for 500 kton year exposure of ical for nh with the true choices of @xmath44 are shown in the left panel of fig .
[ ical - t2k - comp ] .
similar results hold true for inverted hierarchy as the true hierarchy .
it can be seen that the extension of the energy range for analysis and constraining the @xmath45 flux ratio in ical result in an improved sensitivity to the precision on both parameters . the projected ical precision on @xmath3 is better than or comparable to the current t2k precision as can be seen from the right panel of fig .
[ ical - t2k - comp ] , where the 90% cl contour is compared to the results from t2k @xcite .
a comparison of these projected results for ical ( nh ) with the current results from minos @xcite and t2k @xcite are shown in fig .
[ cl - pre ] .
it must be remembered , though , that these experiments are already taking data while ical is yet to be constructed ! plane .
the red dot shows the true choice of parameters for the ical analysis , @xmath46 with nh .
( right ) comparison of the ical projected 10 year sensitivity with the current t2k result at 90% cl .
the ical contour is shifted to match the best - fit value of @xmath14 from t2k @xcite .
, title="fig:",scaledwidth=45.0% ] plane .
the red dot shows the true choice of parameters for the ical analysis , @xmath46 with nh .
( right ) comparison of the ical projected 10 year sensitivity with the current t2k result at 90% cl .
the ical contour is shifted to match the best - fit value of @xmath14 from t2k @xcite .
, title="fig:",scaledwidth=45.0% ] contours at 90% c.l . in the @xmath3@xmath47 plane for 500 kton
yrs exposure of ical ( 3d analysis with @xmath43 = 0.525 gev and the inclusion of the 11th pull ) .
the 90% confidence contours from minos @xcite and t2k @xcite are also shown for comparison .
the red dot shows the true choice of parameters for the ical analysis , @xmath48.,scaledwidth=50.0% ] ical with its magnetisability is an exclusive mass hierarchy machine .
most importantly , its ability to discriminate the normal and inverted mass ordering is _ independent _ of the cp phase @xcite .
the ability of ical to distinguish the correct mass ordering in the 23 sector is given by : @xmath49 where @xmath50 is the minimum @xmath2 calculated using the false ordering , while allowing @xmath15 , @xmath16 and @xmath51 to vary over the @xmath35 range given in table [ osc - par-3sig ] and @xmath52 is the minimum @xmath2 assuming the true mass ordering .
the plot of @xmath53 as a function of the number of years of exposure is shown in fig .
[ hie - plots ] both when the true ordering is taken to be normal ( no ) as well as inverted ( io ) . as a function of exposure time in years .
the extension of observed muon energy range to 0.525 gev increases the sensitivities in the case of both 2d and 3d .
the extra pull improves the hierarchy sensitivity when the true ordering is inverted.,title="fig:",scaledwidth=42.0% ] as a function of exposure time in years .
the extension of observed muon energy range to 0.525 gev increases the sensitivities in the case of both 2d and 3d .
the extra pull improves the hierarchy sensitivity when the true ordering is inverted.,title="fig:",scaledwidth=42.0% ] * normal ordering * : the muon - only analysis ( 2d ) with 10 pulls only gives a @xmath53 of @xmath54 5.2 for 10 years of ical exposure , better than the earlier result of @xmath54 4.6 , a 13% increase in the hierarchy sensitivity .
again the addition of the 11th pull does not improve @xmath53 in both energy ranges in the 2d analysis .
the addition of hadron energy as the third observable increases the @xmath53 to @xmath55 , for an exposure time of 10 years , an improvement over the earlier value of 7.7 .
again the addition of the 11th pull has no effect on the hierarchy sensitivity when normal ordering is taken to be the true ordering . * inverted
ordering * : while trends are similar to that with normal order as the true order , the inclusion of the 11th pull has _ significant impact _ on mass hierarchy sensitivity when the true ordering is _
inverted_. in fact , the best sensitivity ( @xmath43 = 0.525 gev , with 11 pulls and 3d binning ) is better with the inverted than with the normal hierarchy ( by @xmath54 16% ) .
the reason for this effect with the 11@xmath56 pull is discussed in the next section .
it should be noted that the values of @xmath53 are lower than those reported in ref .
@xcite , for the same exposure time .
this is due to the fact that the earlier analysis used the input value @xmath57 while our analysis uses the current best fit value @xcite @xmath58 . given that the best fit of @xmath16 has reduced further to @xmath59 @xcite it is even more important to perform the analysis in as wide a kinematic range ( energy and direction ) as possible to get the best possible hierarchy discrimination .
the sensitivity to mass ordering is also known to depend on the true value of @xmath15 ; it is higher for larger @xmath15 as can be seen from fig .
[ fig : hie-23 ] . while ical has better sensitivity to the inverted ordering when @xmath15 is in the first octant , the reverse is true when it is in the second octant .
this is due to the dominance of the @xmath60 term in eq .
[ toteve ] arising from the survived @xmath0s compared to the @xmath61 term from oscillated @xmath18s , and the nature of its dependence on @xmath15 .
in fact , the results would also be vastly improved by `` switching off '' the @xmath61 term since their dependence on the oscillation parameters ( especially @xmath15 ) are practically the opposite of each other @xcite .
this is not possible with atmospheric neutrinos where nature provides both flavours , but is possible with neutrino beams . in the latter case
, however , the fact that there is a single base - line which results in a significant dependence on and correlation with the cp violating phase @xmath36 and complicates the analysis , as with minos @xcite , t2k @xcite , lbne @xcite or no@xmath62a @xcite . as a function of exposure time in years .
it can be seen that the sensitivity to mass hierarchy increases linearly with the true value of @xmath15 when it is varied over the @xmath35 range of the parameter.,title="fig:",scaledwidth=47.0% ] as a function of exposure time in years .
it can be seen that the sensitivity to mass hierarchy increases linearly with the true value of @xmath15 when it is varied over the @xmath35 range of the parameter.,title="fig:",scaledwidth=47.0% ] to summarise , the sensitivity to the mass ordering in the 23 sector improves with the addition of higher energy bins in the analysis , while constraining the @xmath63 flux ratio improves the sensitivity only when the true ordering is inverted . the ical s ability to determine this mass ordering is significant owing to its magnetisability and its 50 kton mass .
improvement in energy resolutions will further improve the detector s sensitivity to this parameter .
also , in ical , the mass ordering can be determined independent of the cp violating phase @xmath36 because of the range of baselines involved in atmospheric neutrinos @xcite ; this is in contrast to beam / short - base - line experiments where there is a non - trivial sensitivity to the 23 mass ordering depending on the true value of the cp phase .
hence ical will be important in the determination of this mass ordering , and any amount of improvement in determining this parameter is noteworthy .
the 11@xmath65 pull accounts for the fact that the _ ratios _ of the @xmath0 and @xmath1 fluxes are better known ( to within 5% ) than the absolute fluxes themselves @xcite .
this is implemented by using a pull @xmath66 , which contributes _ with the opposite sign _ for neutrino and anti - neutrino events , as can be seen from eq .
[ pi6-xi6 ] .
it is seen that the inclusion of the 11th pull is most visible in the determination of @xmath15 which becomes more constrained when this pull is included .
one way to understand this is to re - bin the events in a single variable @xmath67 ( of the final state muon ) and consider the effect of just this pull .
[ fig : pull ] shows the effect of @xmath15 on both the neutrino and anti - neutrino events .
the thick solid line is the `` data '' corresponding to @xmath68 while the thin solid theory " line corresponds to a fit with @xmath69 and without any pull . in both cases , reducing @xmath15 from the true value increases the event rate _ in every bin _ ( the opposite will hold with the inverted hierarchy ; here the normal hierarchy is shown ) .
note that the down - going events are not shown here . and
@xmath1 events shown as a function of @xmath70 of the final state muon .
the red ( thick solid ) lines correspond to the `` data '' and the black ( thin solid ) ones to the theory " with @xmath71 ( and other parameters held fixed ) .
the green ( dashed ) lines ( practically overlapping the data ) and the blue ( dot - dashed ) lines mimic the effect of 10th pull and inclusion of the 11th pull respectively.,scaledwidth=70.0% ] the dot - dashed and dashed curves corresponding to the labels 11- and 10-pulls show the effect of changing the ( single ) normalisation of the theory with and without the 11@xmath65 pull .
the overall normalisation of the events in the 10-pull case can be _ independently _ varied for neutrinos and anti - neutrinos ( decreased by 4% and 3% in figure ) to improve the agreement of the @xmath72 theory line to the data , resulting in smaller overall @xmath2 in this fit . on the other hand , a 2.5% decrease in the normalisation of neutrino - induced events in the 11-pull case
is accompanied by a 2.5% _ increase _ in the anti - neutrino case , so that the agreement with the neutrino data becomes better , but that with the anti - neutrino data becomes worse .
( of course , it can be applied vice versa , but the smaller @xmath2 is obtained with this choice since there are about twice as many neutrino events as anti - neutrino ones due to the smaller cross section of the latter . )
hence it is not possible to improve the agreement of the @xmath72 theory with data by tuning the normalisation in the analysis when including this pull ; this results in a larger @xmath2 compared to the 10-pull case where the normalisations of neutrino- and anti - neutrino - induced events can be independently varied .
this gives rise to the tighter constraints on @xmath15 when the 11th pull is added .
we note that only a detector like ical that is capable of charge identification can successfully implement this pull as a constraint .
it can also be seen from fig .
[ fig : pull ] that there is greater sensitivity to @xmath15 in the neutrino rather than in the anti - neutrino sector .
the reverse is true with inverted mass ordering . in the determination of the sensitivity to the mass ordering
, the minimum @xmath2 with the false ordering is found .
that means , when the true ordering is inverted , the `` theory '' is obtained using the normal ordering , where there is greater sensitivity to @xmath15 as just mentioned .
when the 11th pull is included , therefore , the discrepancy between theory and data can not be achieved by changing @xmath15 and hence there is more sensitivity to determination of the mass ordering when the true ordering is inverted , provided @xmath15 is in the first octant .
the reverse is true when @xmath15 is in the second octant , as can be seen from fig .
[ fig : hie-23 ] . in addition
, it can be seen from fig .
[ fig : pull ] that there is sensitivity to oscillation parameters near and beyond @xmath73 .
this is precisely the region that is included when the range of @xmath43 is extended from 1 gev down to 0.5 gev .
similarly , although smaller , sensitivity to the parameters is also seen for @xmath743 , which corresponds to the extension in the higher energy end from 11 to 25 gev .
this paper contains a simulation study of the physics potential of the proposed 50 kton magnetised iron calorimeter detector ( ical ) at ino which aims to probe neutrino oscillation parameters by observing atmospheric neutrino oscillations and studying their earth matter effects as they propagate through the earth .
this will be done by detecting ( mainly ) the charged current interactions of @xmath0 and @xmath1 in the detector by means of the final state muons .
the detector , which is optimised for the detection of muons in the gev energy range , will have a magnetic field which will enable the distinction of @xmath0 and @xmath1 events by identifying the charge of the muon in the final state , thus making ical an excellent detector to determine the neutrino mass ordering .
not only this , the magnetic field helps to improve the precision measurement of the mixing angle @xmath15 ( and @xmath75 and the mass ordering as well ) .
the main themes of our study were the effects of extending the observed muon energy range to 0.525 gev ( from 111 gev used in earlier studies ) and that of a constraint on the @xmath0@xmath1 flux ratio on the sensitivity of a 50 kton ical to neutrino oscillation parameters in the 23 sector .
the second and the aspect which is found to have the biggest impact so far on oscillation sensitivity arises from two facts ; one that ical detects atmospheric neutrinos and the other that this massive detector will be magnetised .
in particular , we show that the relatively small uncertainties on the atmospheric neutrino - antineutrino flux _ ratios _ act as a constraint in analyses where the neutrino and antineutrino events can be separated .
this is true in a magnetised detector such as ical where the magnetic field distinguishes muons and anti - muons produced in charged current interactions of neutrinos and antineutrinos respectively with the detector material .
the presence of the magnetic field enables ical to distinguish between neutrino and anti - neutrino events ; hence , inclusion of the uncertainty on the @xmath76 flux ratio as an additional pull translates to a constraint on this ratio which in turn significantly improves the precision with which @xmath3 can be determined .
such a constraint is applicable for all magnetised detectors which have good charge identification capabilities and the impact of this constraint has been shown for the first time in this paper . as a consequence , our simulation studies show that , with 10 years of data taking , ical will not only be able to determine @xmath77 with good precision ( as expected ) but can also pin down @xmath3 to a precision better than the current limits set by t2k .
the precision that can be achieved on these parameters in about 10 years running is about 2.5% and 9% for @xmath78 and @xmath3 respectively ; see details in tables [ pre - tab - stt23 ] and [ pre - tab - dmeff2 ] .
the studies presented here assume that there is perfect separation between different types of charged current ( cc ) and neutral current ( nc ) events .
the event separation efficiency will affect the results of the analysis since they will determine the actual number of events in each bin apart from the contamination from nc oscillation - independent events .
however the inclusion of this consideration is beyond the scope of this paper .
preliminary studies @xcite show that certain selection criteria can be applied so that an event sample which comprises more than 95% cc muon events can be obtained .
the criterion results in cutting out events with @xmath79 gev ; this has determined largely the range ( lower limit ) of muon energy analysed in this paper .
also , improvements in the reconstruction efficiencies and resolutions ( that have been studied with geant-4 simulations of the detector ) as well as possible changes in detector geometry can all alter the results of this analysis .
note that the current studies were all done with the atmospheric neutrino fluxes computed at the super kamiokande site @xcite .
the fluxes at theni where ical is proposed to be built are slightly different and are smaller at energies less than 10 gev @xcite ; this will also impact the analysis .
even within these limitations , it appears that the physics results of ical will have the capability to impact global fits to neutrino data and thus any new analysis will open a window to understanding the neutrino oscillation parameters better on the whole and the atmospheric neutrino fluxes themselves , in particular .
as discussed in section [ nevt ] , the analyses in the previous sections were done by taking a 1000 year sample of charged current muon neutrino events and scaling it down to the required number of years for comparison with `` data '' .
it is important to reduce `` theory '' fluctuations so as to obtain a genuine result from the analysis .
the effect of taking different years of exposure as theory samples , to be scaled to 10 years on precision measurements of @xmath15 and @xmath75 , is shown in fig .
[ chi2-fluct ] for arbitrarily chosen values of @xmath80 and @xmath81 ev@xmath82 ( @xmath83 ev@xmath82 ) .
it can be seen that a smaller sample has more fluctuations and hence yields a larger value of @xmath84 for a given parameter thus giving too good a precision on the oscillation parameter which is false .
the larger sample takes care of this by reducing the ( @xmath85 ) fluctuations in the theory itself .
the @xmath2 stabilises at a point when the sample is fairly large and this is the number of years of exposure to be taken and scaled down for the analysis . it can be seen that use of 1000 year sample yields fairly stable results and thus the analyses have made use of such a sample of charged current muon neutrino events .
we note however that the results are much more sensitive to the sample size in the case of @xmath86 than for @xmath15 . vs the number of years of sample theory to be scaled to 10 years for @xmath87 ( left ) and @xmath88 ( right ) .
it can be seen that the scaling of only 100 years of sample size gives too high a @xmath2 which is more prominent in the case of @xmath89 . as the number of years of sample to be scaled increases
, the value of @xmath2 also comes down as well as saturates for large enough sample sizes.,title="fig:",scaledwidth=45.0% ] vs the number of years of sample theory to be scaled to 10 years for @xmath87 ( left ) and @xmath88 ( right ) .
it can be seen that the scaling of only 100 years of sample size gives too high a @xmath2 which is more prominent in the case of @xmath89 . as the number of years of sample to be scaled increases
, the value of @xmath2 also comes down as well as saturates for large enough sample sizes.,title="fig:",scaledwidth=45.0% ]
we thank prof .
nita sinha , imsc , for many discussions .
we also thank meghna k.k . for the muon resolution table in the lower muon energy range .
we thank the ino internal referees for their valuable and insightful comments on the draft .
lsm thanks dae india and dst india , for funding this research and is thankful for the excellent computing facilities and support from the computing section of imsc , chennai , which made the extensive computing for this analysis possible .
m. honda , t. kajita , k. kasahara , s. midorikawa , and t. sanuki , _ calculation of atmospheric neutrino flux using the interaction model calibrated with atmospheric muon data _ , _ phys .
d _ * 75 * ( 2007 ) 043006 . | we determine the sensitivity to neutrino oscillation parameters from a study of atmospheric neutrinos in a magnetised detector such as the ical at the proposed india - based neutrino observatory .
in such a detector that can _ separately _ count @xmath0 and @xmath1-induced events , the relatively smaller ( about 5% ) uncertainties on the neutrino anti - neutrino flux ratios translate to a constraint in the @xmath2 analysis that results in a significant improvement in the precision with which neutrino oscillation parameters such as @xmath3 can be determined .
such an effect is unique to all magnetisable detectors and constitutes a great advantage in determining neutrino oscillation parameters using such detectors .
such a study has been performed for the first time here . along with an increase in the kinematic range compared to earlier analyses ,
this results in sensitivities to oscillation parameters in the 23 sector that are comparable to or better than those from accelerator experiments where the fluxes are significantly higher .
for example , the @xmath4 precisions on @xmath3 and @xmath5 achievable for 500 kton yr exposure of ical are @xmath6 and @xmath7 respectively for both normal and inverted hierarchies .
the mass hierarchy sensitivity achievable with this combination when the true hierarchy is normal ( inverted ) for the same exposure is @xmath8 ( @xmath9 ) . |
extremely large size waves ( commonly known as freak , rogue or giant waves ) are very common in the open sea or ocean and they pose major hazard to mariners . as early as 1976 , peregrine @xcite suggested that in the region of oceans where there is a strong current present , freak waves can form when action is conncentrated by reflection into a caustic region .
a variable current acts analogously to filamentation instability in laser - plasma interactions @xcite .
freak waves are very steep and is a nonlinear phenomena , hence they can not be represented and described by a linear water wave theory .
zakharov @xcite in 2009 has noted that in the last stage of their evolution , their steepness becomes ` infinite ' , thereby forming a ` wall of water ' .
however , before such an instant in time , the steepness is higher than one for the limiting stokes wave and before breaking the wave crest reaches three to four ( sometimes even more ) times higher than the crests of neighboring waves .
the freak wave is preceded by a deep trough appearing as a ` hole in the sea ' . on the other hand ,
a characteristic life time of a freak wave is short , typically ten of wave periods or so .
for example , if the wave period is fifteen seconds , then their life time is just few minutes .
freak wave appears almost instantly from a relatively calm sea .
it is , therefore , easy to appreciate that such peculiar features of freak waves can not be explained by means of a linear theory .
even the focusing of ocean waves is a preconditions for formation of such waves .
it is now quite common to associate appearance of freak waves with the modulation instability of stokes waves .
this instability ( known as the benjamin feir instability ) was first discovered by lighthill @xcite and the detail of theory was developed independently by benjamin and feir @xcite and by zakharov @xcite .
zakharov showed slowly modulated weakly nonlinear stokes wave can be described by nonlinear shrdinger equation ( nlse ) and that this equation is integrable @xcite and is just the first term in the hierarchy of envelope equations describing packets of surface gravity waves .
the second term in this hierarchy was calculated by dysthe @xcite . since the pioneer work of smith @xcite , many researchers attempted ( both theoretically or numerically ) to explain the freak wave formation by nlse . among diverse results obtained by them
there is one important common observation which has been made by all , and that is , nonlinear development of modulational instability leads to concentration of wave energy in a small spatial region .
this marks the possibility for formation of freak wave .
modulation instability leads to decomposition of initially homogeneous stokes wave into a system of envelope solitons , or more strictly quasi - solitons @xcite .
this state can be called ` solitonic turbulence ' , or ` quasisolitonic turbulence ' . in this letter , we consider the problem of a single solition in a homogeneous media , being subjected to modulational instability which eventually leads to formation of a system of soliton .
we will show that the supercritical instability leads to maximum formation of soliton , concentrated in a small region . in going through subcritical instability the solitons coagulate to early stages of supercritical instability .
moreover , we investigate the full dynamics of nonlinearly interacting deep water waves subjected to modulational / filamentation instabilities , and we find that random perturbations can grow to form inherently nonlinear water wave structures , the so called freak waves , through the nonlinear interaction between two coupled water waves .
the latter should be of interest for explaining recent observations in water wave dynamics .
in a pioneering work , a theory for the modulational instability of a pair of two - dimensional nonlinearly coupled water waves in deep water , as well as the formation and dynamics of localized freak wave packets was presented @xcite .
the two wave packets were investigated in the context of nonlinear optics by @xcite , in bise - einsten condensates by @xcite , in transmission by @xcite , and in plasmas by many other authors @xcite .
following @xcite , we consider the two - dimensional cnls equations in the following form @xmath0 and @xmath1 where @xmath2 and @xmath3 are the amplitudes of the slowly varying wave envelopes .
the @xmath4 and @xmath5 components of the group velocity are given respectively by @xmath6 and the group velocity dispersion coefficients are @xmath7 also , the nonlinearity coefficients ( as in @xcite ) are given by @xmath8 and @xmath9 here @xmath10 and @xmath11 are wavenumbers and @xmath12 is the wave frequency .
they are related by @xmath13 ( the dispersion relation for deep water waves @xcite ) with @xmath14 the acceleration due to gravity and @xmath15 the wavenumber norm given by @xmath16 . for detail description of the formulation of the problem
, we refer to the original works @xcite .
the nonlinear strongly coupled system of equations ( [ cnlse1 ] ) and ( [ cnlse2 ] ) will be computed using a fast numerical algorithm based on the spectral method @xcite which is explained below .
we first noticed that by letting @xmath17 and @xmath18 , the system ( [ cnlse1 ] ) and ( [ cnlse2 ] ) becomes symmetric , obtained from ( [ cnlse1 ] ) and ( [ cnlse2 ] ) @xmath19 where @xmath20\ ] ] then , we reduce the above system of pdes ( [ eq1])([eq2 ] ) into a system of odes using the fourier transform of @xmath21 which is defined by @xmath22 with the corresponding inverse @xmath23 the function @xmath24 can be interpreted as the amplitude density of @xmath25 for wavenumbers @xmath26 , @xmath27 .
now , we take the fourier transform of both ( [ eq1 ] ) and ( [ eq2 ] ) as @xmath28 @xmath29 letting @xmath30 and @xmath31 ( [ eq6 ] ) and ( [ eq7 ] ) can be written in the matrix form as @xmath32 computing the eigenvalues and eigenvectors the solution to ( [ eq8 ] ) can be written as @xmath33 with @xmath34 and @xmath35 .
we exploit the symmetry of the nonlinear function @xmath36 from ( [ eq5 ] ) in developing a numerical procedure to solve the system of odes ( [ eq8 ] ) .
we discretize the spatial domain @xmath37 \times [ -l/2,l/2]$ ] into @xmath38 uniformly spaced grid points @xmath39 with @xmath40 , @xmath41 even , and @xmath42 the length of the rectangular mesh @xmath43 .
given @xmath44 , we define the 2d discrete fourier transform ( 2dft ) of @xmath25 as @xmath45 and its inverse 2dft as @xmath46 in equation ( [ eq10 ] ) and ( [ eq11 ] ) the wavenumbers @xmath26 and @xmath27 , and the spatial indexes @xmath47 and @xmath48 , take only integer values . + we solve the initial value problem of the ode system ( [ eq8 ] ) using the classical fourth order runge - kutta ( rk4 ) method and exact treatment for the linear part @xcite .
given @xmath49 , we discretize the time domain @xmath50 $ ] with equal time steps of size @xmath51 with @xmath52 , and define @xmath53 and @xmath54 . initializing @xmath55 and @xmath56 , we compute the fourier transforms of the nonlinear terms @xmath57 and @xmath58 , and advanced the ode ( [ eq8 ] ) in time with time step @xmath51 using the explicit rk4 for the nonlinear part , together with an exact solution for the linear part as shown in ( [ eq9 ] ) .
the numerical code for the above procedure is implemented in fortran 90 and executed on a linux cluster .
the initial profiles for @xmath2 and @xmath3 were taken as gaussians , @xmath59 in the simulations reported here , we used the parameter values @xmath60 , @xmath61 , @xmath62 , @xmath63 , @xmath64 , @xmath65 , @xmath66 , @xmath67 and a grid of @xmath68 nodes in the computational domain @xmath69\times[-1,1]$ ] with the time step size @xmath70 . for each simulation
we monitor the energies @xmath71 and @xmath72 , calculated as @xmath73 @xmath74 observing a finite energy will reveal stability of a solution . as soon as the solution becomes unstable , the energy diverges . when the solution dissipates the energy approaches to zero .
( left ) and @xmath72 ( right ) in a typical 3000 min .
simulation.,title="fig:",width=230,height=161 ] [ fig1 ]
which are propagating at an angle of @xmath75 .
a low - amplitude noise equal to @xmath76 is added to the initial amplitude in order to enhance the modulation instability.,title="fig:",width=230,height=161 ] [ fig2 ] the results presented in this paper represent a preliminary study on dynamics of interacting nonlinear water waves .
the problem considered here comprises the dynamics of nonlinear interacting water wave packets through solving the coupled system of equations ( [ cnlse1 ] ) and ( [ cnlse2 ] ) .
the results that are shown in fig .
2 are all in dimensional units , where the two interacting waves initially have the amplitude @xmath77 , with ran representing a random low - amplitude noise , equal to @xmath76 , in order to enhance instability .
the results shown represent different time steps ( starting on the left - hand panel and going downwards ) for @xmath78 , @xmath79 , @xmath80 then ( right - hand panel ) @xmath81 , @xmath82 ; the last figure on the right - hand panel is at the same time as that above it but plotted from a different prospective reflecting the maximum growth rate in the @xmath5 direction . for our simulations
we have taken typical data from ocean waves @xcite .
the waves @xmath2 and @xmath3 in fig .
2 , then have the initial amplitudes @xmath83 meters . from these figures ,
we see at @xmath84 ( @xmath85 seconds ) that wave @xmath2 focuses as a localized wave packets with a maximum amplitude of @xmath86 meters .
we remark for considerable period after the initial step , waves @xmath2 and @xmath3 are qualitatively the same ( with @xmath87 ) before the nonlinear wave - wave interactions set in which results to wave break - up . in summary
, we presented a numerical procedure to solve cnls equations describing modulational instabilities of a pair of nonlinearly interacting two - dimensional waves in deep water .
the simulation results of the full dynamical system reveals that even waves that are separately modulationally stable can , when nonlinear interactions are taken into account , give rise to novel behavior such as the formation of large - amplitude coherent wave packets with amplitudes several times the initial waves .
this behavior is quite different from that of a single wave ( the case for the original benjamin - feir instability ) which disintegrates into a wide spectrum of waves .
these results are relevant to the nonlinear instability arising from colliding water waves thereby producing large - amplitude oceanic freak waves . | the coupled cubic nonlinear schrdinger ( cnls ) equations are used to study modulational instabilities of a pair of nonlinearly interacting two - dimensional waves in deep water .
it has been shown that the full dynamics of these interacting waves gives rise to localized large - amplitude wavepackets ( wave focusing ) . in this short letter
we attempt to verify this result numerically using a fourier spectral method for the cnls equations . |
the p - mode oscillation frequencies vary with the activity cycle of the sun @xcite and other solar - like stars @xcite .
however , as observed in the sun , these temporal variations present differences in amplitude , phase , and frequency among individual modes . thus the frequency separations computed will be observed to vary with time .
comparing these frequency separations with those from stellar models will lead to differences in the fitted global parameters ( radius , mass , age ) of the star . generally , when the frequencies are measured we do not know the phase of the activity cycle , unless the star can be measured continuously for long periods of time .
we should then take into account when inferring the stellar parameters , that these values may depend on when the data was obtained .
we analyzed 5202 days of velocity observations collected by the space - based instrument golf onboard _
soho _ spacecraft , covering more than 14 years between 1996 and 2010 with an overall duty cycle of 95.4% @xcite .
this dataset was split into non - independent , contiguous 365-day subseries with 91.25-day overlap and their associated power spectra fitted to extract the mode parameters @xcite using a standard likelihood maximization function ( power spectrum with a @xmath0 with 2 d.o.f .
statistic ) .
the formal uncertainties in each parameter were then derived from the inverse hessian matrix .
the individual frequency separations measured with 365-day time series were obtained with a precision of about 0.07 @xmath1hz .
figure [ fig : data ] ( left panel ) shows the temporal variations of the individual @xmath2 and @xmath3 mode frequencies averaged between 2200 and 3300 @xmath1hz ( the reference values being taken as the average over 19961997 ) .
it is clear that the individual sun - as - a - star p - mode frequencies have different temporal variations @xcite , which are consistent between radial velocity ( golf ) and intensity virgo observations @xcite .
the averaged large ( @xmath4 ) and small ( @xmath5 ) frequency separations also show significant temporal variations with solar activity .
for example , figure [ fig : data ] ( right panel ) shows that the small frequency separation @xmath6 varies from peak - to - peak by about @xmath7hz over the solar cycle , which is presumably consistent with being signatures from surface effects . for a broader perspective
, it is very important to note that the sun is considered a low - activity star , and many solar - type stars could exhibit much larger variations with activity cycle .
in order to determine the global model parameters of the sun @xmath8 , we match the observational data @xmath9 to the observables @xmath10 from stellar models @xcite which are characterized by @xmath8 .
@xmath8 comprises the mass m , age a , initial hydrogen ( or helium ) mass fraction @xmath11 , initial heavy element mass fraction @xmath12 , and the mixing - length parameter @xmath13 ( from the standard mixing - length theory of @xcite ) . the parameters that best describe the data are obtained by minimizing a @xmath0 function : @xmath14 where there are @xmath15 independent observations and @xmath16 is the observational error .
we use the levenberg - marquardt algorithm ( lm ) to minimize eq .
[ eqn : chi2 ] @xcite .
the global objective of this work is to investigate the effect of the shifted frequency values on the determination of @xmath8 for any star with solar - like oscillations .
so we must define a coherent and consistent method to fit the observational data to obtain the set of best - fitting parameters not only for the sun , but for other stars too .
our strategy is to first fit the non - seismic and the average seismic quantities to obtain @xmath8 to a reasonable range , and then using @xmath8 as initial guesses , proceed to use the individual frequency separations ( large and small ) to obtain the final set of best - fitting parameters .
lm is a local minimization method and can therefore be sensitive to initial conditions . to ensure that we obtain a global minimum we search for the best parameters by initializing the minimization with different parameter values .
we set the initial @xmath10 as 1.0 @xmath17 , and the initial a varies between 4 and 6 gyr in steps of 0.5 gyr .
additionally , we hold the parameter @xmath11 fixed during each minimization , and instead repeat the process for three values of @xmath11 : 0.69 , 0.71 , 0.73 .
this gives in total 15 minimizations using the input data at maximum activity and again at minimum activity .
from the 15 best - fitted @xmath8 , we selected those sets where @xmath18 ( 4 fitted parameters , 19 data points to give 14 degrees of freedom at the 1% confidence level ) and where the fitted @xmath19 fell to within 1@xmath16 of 5777 k ( @xmath19 was not a constraint in the second part of the fitting strategy ) .
these sets are represented in fig .
[ fig : models ] , where we show the values of age versus mass on the left panel , and radius versus effective temperature on the right .
the dark / lighter filled circles are the results for fitting during maximum / minimum activity .
[ fig : models ] left panel shows an offset in the fitted @xmath8 between minimum and maximum activity . to quantify this offset , we fitted a linear function to each group of results .
the difference between the fitted ages at opposite activity phases is 0.2 gyr .
alternatively if we take 4.8 gyr as the solar age , the mass is fitted as a 1.0 @xmath17 star at minimum , and 1.03 @xmath17 at maximum .
the right panel of fig .
[ fig : models ] shows the fitted @xmath20 and @xmath19 for these same models ( neither were input constraints ) .
we find that at minimum activity , the sun s @xmath19 is about 50 k cooler for a given @xmath20 , or about 0.5% smaller in @xmath20 for a given @xmath19 .
measuring the oscillation frequencies during different phases of a stellar activity cycle will lead to different values of the individual frequencies .
thus , estimates of the large and small frequency separations will be different depending on the observing period . using 365-day subseries of golf data
, we determined the individual frequency separations to about 0.07 @xmath1hz .
we used stellar models to determine the best - fitting parameters using the observations taken at minimum and maximum activity .
we found that the age of the star is on average 0.2 gyr older using the values from minimum activity , or for a given age , the mass is about 2% larger using the data at maximum activity .
we also found a small discrepancy in the fitted radius and effective temperature . at minimum activity
, the sun is about 50 k cooler for a given fitted radius , or the radius is about 0.5% smaller for a given effective temperature .
although we still need to study whether these differences in fitted values are detectable given the expected uncertainties , we note that @xcite quote an error in the age of the planet - hosting star hat - p-7 of 0.26 gyr , of the order of this detected change .
the sun , however , is considered a low - activity star , and much stronger activity cycles on other solar - type stars have already been detected @xcite . with long time - series data such as those from corot or kepler , we will not only detect stellar activity cycles , but we will obtain very high precision on the seismic data , that the uncertainties in the fitted parameters will be smaller than the fitted changes in these parameters . the authors want to thank catherine renaud for the calibration and preparation of the golf dataset .
the golf instrument onboard soho is a cooperative effort of many individuals , to whom we are indebted .
soho is a project of international collaboration between esa and nasa.this research was in part supported by the european helio- and asteroseismology network ( helas ) , a major international collaboration funded by the european commission s sixth framework programme .
ds acknowledges the support of the grant pnaya2007 - 62650 from the spanish national research plan . this work has been partially supported by the cnes / golf grant at the sap cea - saclay .
9 salabert d , garca r a , pall p l and jimnez - reyes s j 2009 _ a&a _ * 504 * l1 garca r a et al .
2010 _ science _ * 329 * 1032 garca r a et al . 2005 _
a&a _ * 442 * 385 salabert d , chaplin w j , elsworth y , new r and verner g a 2007 _ a&a _ * 463 * 1181 salabert d , garca r a , pall p l and jimnez a 2010 _ j. phys . :
conf . series _
these proceedings christensen - dalsgaard j 2008a _ ap&ss _ * 316 * 113 christensen - dalsgaard j 2008b _ ap&ss _ * 316 * 13 bhm - vitense e 1958 _ za _ * 46 * 108b creevey et al .
2007 _ apj _ * 659 * 616 metcalfe t , creevey o l and christensen - dalsgaard j 2009 apj * 699 * 373 stello et al .
2009 _ apj _ * 700 * 1589 chaplin et al .
2010 _ apjl _ * 713 * 169 mathur et al . 2010
_ a&a _ * 511 * 46 christensen - dalsgaard j et al .
2010 _ apj _ * 713 * 164 | frequency separations used to infer global properties of stars through asteroseismology can change depending on the strength and at what epoch of the stellar cycle the p - mode frequencies are measured . in the sun
these variations have been seen , even though the sun is a low - activity star . in this paper
, we discuss these variations and their impact on the determination of the stellar parameters ( radius , mass and age ) for the sun . using the data from maximum and minimum activity
, we fitted an age for the sun that differs on average by 0.2 gyr : slightly older during minimum activity .
the fitted radius is also lower by about 0.5% for the solar effective temperature during minimum . |
the standard model ( sm ) of particle physics , though it can fit most of the present data , leaves a few fundamental problems unsolved , one of which is the gauge hierarchy problem . since the second revolution of string theories , a crop of models with extra dimensions have been proposed to solve various theoretical problems . in an attractive model of large extra dimensions or tev quantum gravity ( add model ) @xcite , the fundamental planck scale can be as low as a few tev .
this is made possible by localizing the sm particles on a brane ( using the idea of d - branes in type i or ii string theory ) , while gravity is free to propagate in all dimensions .
the observed planck scale ( @xmath5 gev ) is then a derived quantity .
extensive phenomenology studies have been carried out in recent years .
signatures for the add model can be divided into two catergories : sub - planckian and trans - planckian .
the former is the one that was studied extensively , while the latter just recently received more attention , especially black hole production in hadronic collisions .
a black hole ( bh ) has been illusive for decades , as we can not directly measure any properties of it , not to mention the production of black holes in any terrestrial experiments .
this is due to the fact that in order to produce black holes in collider experiments one needs a center - of - mass energy above the planck scale ( @xmath6 gev ) , which is obviously inaccessible at the moment . in models of large extra dimensions ,
the properties of black holes are modified and interesting signatures emerge @xcite . the fact that the fundamental planck scale is as low as tev also opens up an interesting possibility of producing a large number of black holes at collider experiments ( e.g. lhc ) @xcite .
reference @xcite showed that a bh localized on a brane will radiate mainly in the brane , instead of radiating into the kaluza - klein states of gravitons of the bulk . in this case , the bh so produced will decay mainly into the sm particles , which can then be detected in the detector .
this opportunity has enabled investigation of the properties of bh at terrestrial collider experiments .
there have been a number of such studies @xcite at hadronic colliders .
a typical signature of the bh decay is a high multiplicity , isothermal event , very much like a spherical fireball . " on the other hand , bh production has also been studied in cosmic ray experiments @xcite .
the ultrahigh energy cosmic rays ( uhecr ) serve as a very energetic beam hitting on the atmosphere as the target .
the primary ingredient of uhecr is probably protons or light nuclei , or could be photons or even neutrinos .
if it is made up of protons , it could produce a large number of bh s up at the top of the atmosphere , producing a giant air shower .
if it is neutrinos , it could produce very interesting horizontal air showers @xcite , or black holes within a km - sized neutrino telescope @xcite .
an important quantity of a bh is its entropy @xmath7 . to fulfill the thermodynamical description , a bh requires a large entropy of order of 25 @xcite .
such an entropy requirement implies that the bh mass must be at least five times the fundamental planck scale @xcite .
this mass requirement makes the bh production not as large as previously calculated in a number of works @xcite , first pointed out in ref . @xcite .
in addition , the signature of large multiplicity decay of a bh can only happen when the entropy is large . even taking into account this mass requirement , the event rate is still large enough for detection . on the other hand , there were arguments from voloshin that the cross section should be multiplied by an exponential factor @xmath8 @xcite ( we shall get back to this point later in sec .
however , this suppression factor becomes too severe for the production rate to be interesting , contrary to the conclusion of ref .
@xcite , because of the large entropy requirement .
there have been continuous theoretical efforts to calculate the production and decay of bh s in particle collisions @xcite .
other interesting trans - planckian phenomena include string balls @xcite , @xmath0-branes @xcite , and tev string behaviors @xcite . dimopoulos and
emparan @xcite pointed out that when a bh reaches a minimum mass , it transits into a state of highly excited and jagged strings a string ball ( sb ) .
the transition point is at @xmath9 where @xmath1 is the string scale and @xmath3 is the string coupling .
naively , sb s are stringy progenitors of bh s .
the bh correspondence principle states that properties of a bh with a mass @xmath10 match those of a string ball with @xmath11 .
we can then equate the production cross sections of sb and bh at the transition point .
in fact , we shall use this argument to write down the cross section for the sb at the transition point .
the existence of string balls could be argued from the string point of view .
when the energy of the scattering reaches the string scale , the scattering of particles is no longer described by point - particle scattering but replaced by string - string scattering .
as the energy goes further up , the strings become highly excited , jagged and entangled string states , and become like a string ball .
when the energy reaches the transition point , it turns into a bh .
previously , in the discussion of a bh , we mentioned a large entropy requirement on the bh in order for the object to be a bh .
such a large mass requirement makes the production cross section smaller than previously thought . here in the case of sb s ,
the mass requirement is substantially lower , thus the production rate is significantly higher .
hence , an sb is more interesting in the experimental point of view if it decays with a distinct signature .
dimopoulos and emparan @xcite argued that the decay of a sb is similar to that of a bh , i.e. , a high multiplicity decay into the sm particles , though in some intermediate stages the sb decays more likely into bulk particles .
another interesting trans - planckian object is the @xmath0-brane .
a bh can be considered a @xmath12-brane . in particle collisions , if one considers bh production , one should also consider @xmath0-brane production .
in fact , the properties of @xmath0-branes reduce to those of a bh in the limit @xmath13 . in extra dimension models ,
in which there are large extra dimensions and small extra dimensions of the size of the planck length , let a @xmath0-brane wrap on @xmath14 small and @xmath15 large dimensions .
it was found @xcite that the production of @xmath0-branes is comparable to bh s only when @xmath16 , i.e. , the @xmath0-brane wraps entirely on the small dimensions only . if @xmath17 , the production of @xmath0-branes would be suppressed by powers of @xmath18 , where @xmath19 is the fundamental scale of the @xmath20 dimensions .
therefore , here we only consider the case in which @xmath16 .
the decay of @xmath0 branes is not well understood .
one interesting possibility is cascade into branes of lower dimensions until they reach the dimension of zero . whether the zero brane is stable depends on the model .
another possibility is the decay into brane and bulk particles , thus experimentally the decay can be observed . or it can be a combination of cascade into lower - dimensional branes and direct decays . in this work , we study the production rates of the bh s , sb s , and @xmath0-branes in hadronic collisions , with emphasis on the lhc and the vlhc .
the organization is as follows . in the next section ,
we briefly describe the relation between the fundamental planck scale and the string scale . in sec .
iii , we describe the production of bh s , sb s , and @xmath0-branes . in sec .
iv , we show our numerical results for the lhc and vlhc .
we discuss the decays in sec .
v and conclude in sec .
first let us address more clearly the configuration of the space - time .
let there be @xmath21 total extra dimensions with @xmath22 small extra dimensions and @xmath23 large extra dimensions .
when we say small extra dimensions , we mean the size is of order of @xmath24 , the fundamental planck scale .
the observed 4d planck scale @xmath25 is then a derived quantity given by @xcite @xmath26 where @xmath27 and @xmath28 are the volumes of the extra @xmath22 and @xmath23 dimensions , respectively , given by @xmath29 where we have expressed the lengths @xmath30 in units of planckian length @xmath24 . where @xmath31 is the radius of the torus .
] suppose the small extra dimension has the size of @xmath32 , i.e. , @xmath33 , then @xmath34 the fundamental planck scale @xmath19 is lowered to the tev range if the size @xmath35 is taken to be very large , of order @xmath36 .
the relation of the observed planck scale to the string scale @xmath1 is given by @xcite @xmath37 where we again take the small extra dimensions of size @xmath32 . from eqs .
( [ nm ] ) and ( [ nmm ] ) we can relate the string scale with the fundamental planck scale as @xmath38 where we take the proportional constant of order @xmath39 , which depends on different compactification configurations .
we shall also use the more conventional definition of the fundamental planck scale @xmath4 related to @xmath19 by @xmath40 where @xmath41 is the gravitational constant in @xmath42 dimensions ( used in the einstein equation : @xmath43 ) .
then @xmath4 is related to the string scale @xmath1 as @xmath44 for a constant @xmath45 of order @xmath39 . in the next section
we shall simply use @xmath46 for discussions and cross - section calculations .
a black hole is characterized by its schwarzschild radius @xmath47 , which depends on the mass @xmath48 of the bh . a simplified picture for bh production is as follows . when the colliding partons have a center - of - mass ( c.o.m .
) energy above some thresholds of order of the planck mass and the _ impact parameter _
less than the schwarzschild radius @xmath47 , a bh is formed and almost at rest in the c.o.m .
the bh so produced will decay thermally ( regardless of the incoming particles ) and thus isotropically in that frame .
this possibility was first investigated for the lhc in refs .
@xcite . in refs .
@xcite , black hole production in hadronic collisions is calculated in @xmath49 subprocesses : @xmath50 , where @xmath51 are incoming partons . the black hole so produced is either at rest or traveling along the beam pipe such that its decay products ( of high multiplicity ) have a zero net transverse momentum ( @xmath52 ) .
giddings and thomas @xcite and dimopoulos and landsberg @xcite demonstrated that a bh so produced will decay with a high multiplicity . in ref .
@xcite , we pointed out the `` @xmath53 '' subprocesses , such that the bh is produced with a large @xmath52 before it decays .
the `` @xmath53 '' subprocesses can be formed when the c.o.m .
energy of the colliding particles is larger than the bh mass ; the excess energy will be radiated as other sm particles . ) , and everything , including the `` extra partons , '' will be hidden behind the horizon .
thus , the entire energy is contained within the bh , which is considered a quantum state rather than a particle . in this picture
, there is no transverse momentum given to the bh , unless by the initial - state radiation of the incoming partons .
this was already studied by a monte carlo approach in ref .
] in such subprocesses , the `` others '' are just the ordinary sm particles and usually of much lower multiplicity than the decay products of the bh . therefore , the signature is as follows : on one side of the event there are particles of high multiplicity ( from the decay of the bh ) , the total @xmath52 of which is balanced by a much lower number of particles on the other side .
such a signature is very clean and should have very few backgrounds .
the next natural question to ask is how large the event rate is .
collider phenomenology is only possible if the event rate is large enough , especially if we want to study bh properties .
it was first pointed out in ref .
@xcite that the bh production rate is so enormous that the lhc is in fact a bh factory .
however , this has not taken into account the entropy factor of the bh .
it was shown in ref . @xcite
that with a large entropy requirement ( @xmath54 ) the bh production rate decreases substantially , but still affords enough events for detection .
for example , the production cross section is as high as 1 pb for @xmath55 tev and @xmath56 with @xmath57 at the lhc .
this implies @xmath58 events with an integrated luminosity of 100 fb@xmath59 . of course , if we relaxed this entropy constraint , the production cross sections would be increased tremendously , but the cross sections have to be interpreted with care because of the presence of large string effects in this regime .
the schwarzschild radius @xmath47 of a bh of mass @xmath48 in @xmath20 dimensions is given by @xcite @xmath60 where @xmath61 is introduced for convenience and @xmath4 is the fundamental planck scale in the model of large extra dimensions already defined in eq .
( [ m*-md ] ) .
the radius @xmath47 is much smaller than the size of the extra dimensions .
bh production is expected when the colliding partons with a center - of - mass energy @xmath62 pass within a distance less than @xmath47 .
a black hole of mass @xmath48 is formed and the rest of the energy , if there is any , is radiated as ordinary sm particles . this semiclassical argument calls for a geometric approximation for the cross section for producing a bh of mass @xmath48 as @xmath63 in the @xmath49 subprocess , the c.o.m .
energy of the colliding partons is just the same as the mass of the bh , i.e. , @xmath64 , which implies a subprocess cross section @xmath65 on the other hand , for the @xmath66 subprocesses the subprocess cross section is @xmath67 another important quantity that characterizes a bh is its entropy given by @xcite @xmath68 the variation of @xmath7 versus the ratio @xmath69 is shown in fig .
[ sbh ] . to ensure the validity of the above classical description of a bh @xcite ,
the entropy must be sufficiently large , of order 25 or so . from the figure we can see that when @xmath70 , the entropy @xmath71 .
therefore , to avoid getting into the nonperturbative regime of the bh and to ensure the validity of the semiclassical formula , we restrict the mass of the bh to be @xmath72 , where @xmath73 is of order 5 .
voloshin @xcite pointed out that the semiclassical argument for the bh production cross section is not given by the geometrical cross - section area , but , instead , suppressed by an exponential factor , @xmath74^{\frac{n+2}{n+1 } } \exp \left ( - \frac{s_{\rm bh}}{n+1 } \right ) \;.\ ] ] there are , however , counter arguments @xcite that the simple geometric formula should be valid . in ref .
@xcite , we have considered both forms of cross sections : the naive @xmath75 and the @xmath75 multiplied with the exponential factor of eq .
( [ supp ] ) .
the suppression factor renders the cross section too small for detection , because the exponential factor contains the entropy @xmath7 , which has to be sufficient large ( e.g. , @xmath76 ) to define a black hole .
the suppression is more than two orders of magnitude and , therefore , we shall not be concerned with this suppression factor anymore .
this seems to contradict the results of ref .
@xcite , in which rizzo concluded that even when voloshin s suppression factor is included , the cross section is still large enough , because he did not impose a large entropy requirement on the validity of the bh .
if he had done so , he would also have gotten a very large suppression .
nevertheless , careful interpretation is needed if the bh has only a small entropy .
dimopoulos and emparan @xcite pointed out that when a bh reaches a minimum mass , it transits into a state of highly excited and jagged strings , dubbed string ball .
they are the stringy progenitors of bh s and share some properties of bh s , such as large production cross sections at hadronic supercolliders and similar signatures when they decay .
they made an important observation @xcite that the minimum mass @xmath77 ( the transition point ) above which a bh can be treated general - relativistically is @xmath2 , where @xmath1 and @xmath3 are the string scale and the string coupling , respectively .
below this transition point , the configuration is dominated by string balls .
since the mass of a string ball is lower than a bh , the corresponding production cross section is larger than that of a bh .
thus , at the lhc , string ball production may be more important . according to the bh correspondence principle
, the properties of a bh with a mass @xmath10 match those of a string ball with a mass @xmath11 .
therefore , the production cross section of a string ball or a bh should be smoothly joined at @xmath78 , i.e. , @xmath79 the production cross section for string balls with mass between the string scale @xmath1 and @xmath80 grows with @xmath81 until @xmath80 , beyond which , due to unitarity , it should stay constant .
therefore , we can use the bh cross section and match to the string ball cross section at the transition point @xmath2 .
this string ball cross section then stays constant between @xmath80 and @xmath2 .
then below @xmath80 the string ball cross section grows like @xmath82 .
the cross sections for the sb or bh are given by @xmath83 ^ 2 & \frac{m_s}{g_s^2 } \le m_{\rm bh } \\
% \frac{\pi}{m_d^2}\ ; \left ( \frac{m_s / g_s^2}{m_d } \right)^{\frac{2}{n+1 } } \left [ f(n ) \right ] ^2 = \frac { \pi}{m_s^2 } \left [ f(n ) \right ] ^2 & \frac{m_s}{g_s } \le m_{\rm sb } \le \frac{m_s}{g_s^2 } \\ % \frac { \pi g_s^2 m_{\rm sb}^2 } { m_s^4 } \left [ f(n ) \right ] ^2 & m_s \ll m_{\rm sb } \le \frac{m_s}{g_s } \end{array } \right . \;,\ ] ] in which we have set @xmath84 as in eq .
( [ ms - md ] ) with @xmath46 . a graphical presentation of these cross sections is shown in fig .
[ sigma - hat ] for @xmath85 . in the next section ,
when we calculate the production cross sections for bh s and sb s , we use the above equation , together with eq .
( [ ms - md ] ) with @xmath46 .
the production then depends on the following parameters : @xmath86 , and @xmath4 .
the @xmath4 can be determined by eq .
( [ ms - md ] ) .
we also require that at the bh - sb transition point , @xmath87 , the mass of the bh is already at @xmath88 ( this ensures that the bh has a sufficiently large entropy @xmath89 @xcite ) .
therefore , the production cross sections depends on @xmath1 and @xmath21 only .
we shall present the results in terms of @xmath4 and @xmath21 for easy comparison with existing literature .
a black hole can be considered a zero - brane . in principle , higher - dimensional objects , e.g. , @xmath0-branes ( @xmath0b ) , can also be formed in particle collisions , in particular when there exist small extra dimensions of the size @xmath90 in addition to the large ones of the size @xmath91 .
it was pointed out by ahn _
_ @xcite that the production cross section of a @xmath0-brane completely wrapped on the small extra dimensions is larger than that of a spherically symmetric black hole .
a similar situation is true in cosmic ray experiments @xcite .
consider an uncharged and static @xmath0-brane with a mass @xmath92 in @xmath93-dimensional space - time ( @xmath22 small planckian size and @xmath23 large size extra dimensions such that @xmath94 ) .
suppose the @xmath0-brane wraps on @xmath95 small extra dimensions and on @xmath96 large extra dimensions .
then the `` radius '' of the @xmath0-brane is @xmath97 where @xmath98 is the volume wrapped by the @xmath0-brane in units of the planckian length .
recall from eq .
( [ nm ] ) , @xmath99 , where @xmath100 and @xmath101 are the lengths of the size of the large and small extra dimensions in units of planckian length ( @xmath90 ) .
then @xmath98 is given by @xmath102 where we have taken @xmath103 .
the function @xmath104 is given by @xmath105^{\frac{1}{1+n - p } } \;.\ ] ] the @xmath106 reduces to the @xmath47 in the limit @xmath107 .
the production cross section of a @xmath0-brane is similar to that of bh s , based on a naive geometric argument @xcite .
when the partons collide with a center - of - mass energy @xmath108 larger than the fundamental planck scale and an impact parameter less than the size of the @xmath0-brane , a @xmath0-brane of mass @xmath109 can be formed .
that is , @xmath110 therefore , the production cross section for a @xmath0-brane is the same as bh s in the limit @xmath107 ( i.e. , a bh can be considered a @xmath12-brane ) . in @xmath111 and @xmath112 processes , the parton - level cross sections are given by similar expressions in eqs .
( [ 2to1 ] ) and ( [ 2to2 ] ) , respectively . in eq .
( [ rp ] ) , we can see that the radius of a @xmath0-brane is suppressed by some powers of the volume @xmath98 wrapped by the @xmath0-brane .
it is then obvious that the production cross section is largest when @xmath98 is minimal , in other words , the @xmath0-brane wraps entirely on the small extra dimensions only , i.e. , @xmath16 .
when @xmath16 , @xmath113 .
we can also compare the production cross section of @xmath0-branes with bh s .
assuming that their masses are the same and the production threshold @xmath114 is the same , the ratio of cross sections is @xmath115 in the above equation , the most severe suppression factor is in the first set of parentheses on the right - hand side . since we are considering physics of tev @xmath19 , the factor @xmath116 .
thus , the only meaningful production of a @xmath0-brane occurs for @xmath16 , and then their production is comparable . in table
[ pbrane ] , we show this ratio for various values of @xmath21 and @xmath0 .
the production of bh s and sb s depends on @xmath117 , but they are related by eq .
( [ ms - md ] ) .
since we also require the transition point ( @xmath2 ) at @xmath118 , we can therefore solve for @xmath1 and @xmath3 for a given pair of @xmath4 and @xmath21 .
we present the results in terms of @xmath4 and @xmath21 .
the minimum mass requirement for the sb is set at @xmath119 .
the production of a @xmath0-brane also depends on @xmath22 and @xmath14 . for an interesting level of event rates
, @xmath14 has to be equal to @xmath0 , i.e. , the @xmath0-brane wraps entirely on small ( of planck length ) extra dimensions .
so after setting all parameters , we are ready to present our numerical results . in fig .
[ total ] , we show the total production cross sections for bh s , sb s , and @xmath0-branes , including the @xmath49 and @xmath120 subprocesses ( when computing the @xmath120 subprocess we require a @xmath52 cut of 500 gev to prevent double counting )
. typically , the @xmath120 subprocess contributes at a level of less than 10% . for the bh , sb , and @xmath0-brane
, we show the results for @xmath121 and @xmath122 .
the results for @xmath123 lie in between .
since we require @xmath124 , their production is only sizable when @xmath125 reaches about 10 tev , unlike the sb , which only requires @xmath126 .
the @xmath0-brane cross section is about a few times larger than the bh , as we have chosen @xmath127 .
string ball production is , on average , two orders of magnitude larger than that of a bh in the energy range between 20 and 60 tev .
below 20 tev ( e.g. , at the lhc ) , the sb cross section is at least three orders of magnitude larger than the bh .
now we particularly look at the production rates at the lhc , operating at @xmath128 tev with a nominal yearly luminosity of 100 fb@xmath59 .
the differential cross sections @xmath129 , where @xmath130 , are shown in fig .
[ dsdm ] , where we have shown the case of @xmath56 and @xmath55 tev . in our scheme ,
@xmath131 tev .
the minimum sb mass starts at @xmath132 tev , while the bh and @xmath0-brane start at @xmath133 tev .
the sb spectrum smoothly joined to the bh spectrum at the transition point @xmath134 .
similarly , the transverse momentum spectra for their production are shown in fig .
[ dsdpt ] . even at a very high @xmath135 tev ,
the cross section is still large enough for detection .
we show the integrated cross sections for the lhc in table [ table - lhc ] , including contributions from @xmath111 and @xmath136 processes ( we imposed a @xmath52 cut of 500 gev in the @xmath136 process ) .
sensitivity information can be drawn from the table .
the event rates for bh and @xmath0-brane production are negligible for @xmath137 tev and only moderate at @xmath138 tev . at @xmath138 tev ,
the number of bh events that can be produced in one year running ( 100 fb@xmath59 ) is about @xmath139 for @xmath140 while the number for @xmath0-brane events is @xmath141 .
therefore , the sensitivity for a detectable signal rate for a bh and a @xmath0-brane is only around 2 tev , if not much larger than 2 tev .
the sb event rate is much higher .
even at @xmath142 tev , the cross section is of order of 30 pb . in table
[ table - lhc ] , we also show the @xmath143 for @xmath144 tev .
roughly , the sensitivity is around 6 tev .
the vlhc ( very large hadron collider ) is another @xmath145 accelerator under discussions @xcite in the snowmass 2001 @xcite .
the preliminary plan is to have an initial stage of about 4060 tev center - of - mass energy , and later an increase up to 200 tev .
the targeted luminosity is @xmath146 . in fig .
[ total-60 ] , we show the total production cross sections for bh s , sb s , and @xmath0-branes for @xmath147 tev and for @xmath121 and @xmath148 . the integrated cross sections for @xmath149 , and @xmath150 tev are shown in table [ table - vlhc ] . for a fixed @xmath4 ,
the cross section obviously increases with @xmath125 .
we choose to show the event rates for different values of @xmath4 such that it roughtly gives an idea about the sensitivity reach at each @xmath125 .
we found that the sensitivity reaches for bh and @xmath0-brane production are roughly between @xmath148 and @xmath151 tev for @xmath152 tev , @xmath153 and @xmath154 tev for @xmath155 tev , @xmath156 and @xmath157 tev for @xmath158 tev , and @xmath159 and @xmath160 tev for @xmath161 tev .
these estimates are rather crude based on the requirement that the number of raw events is @xmath162 .
the main phase of the decay of a bh is via the hawking evaporation .
the evaporation rate is governed by its hawking temperature , given by @xcite @xmath163 which scales inversely with some powers of @xmath48 .
the heavier the bh , the lower is the temperature .
thus , the evaporation rate is slower .
the lifetime of the bh also scales inversely with the hawking temperature as given by @xmath164 from the above equation , it is obvious that the lifetime of a bh becomes much longer in models of large extra dimensions than in the usual @xmath165 theory .
however , the lifetime is still so short that it will decay once it is produced and no displaced vertex can be seen in the detector . for another viewpoint on the bh decay
, please see ref .
@xcite .
an important observation is that the wavelength @xmath166 of the thermal spectrum corresponding to the hawking temperature is larger than the size of the bh .
this implies that the bh evaporates like a point source in @xmath81-waves , therefore it decays equally into brane and bulk modes , and will not see the higher angular momentum states available in the extra dimensions .
since on the brane there are many more particles than in the bulk , the bh decays dominantly into brane modes , i.e. , the sm particles in the setup .
furthermore , the bh evaporates
blindly " into all degrees of freedom . the ratio of the degrees of freedom for gauge bosons , quarks , and leptons is @xmath167 ( the higgs boson is not included ) .
since the @xmath168 and @xmath169 decay with a branching ratio of about 70% into quarks , and the gluon also gives rise to hadronic activities , the final ratio of hadronic to leptonic activities in the bh decay is about @xmath170 @xcite .
another important property of the bh decay is the large number of particles , in accord with the large entropy in eq .
( [ entropy ] ) , in the process of evaporation .
it was shown @xcite that the average multiplicity @xmath171 in the decay of a bh is order of @xmath172 for @xmath48 being a few times @xmath4 for @xmath173 .
since we are considering the bh that has an entropy of order 25 or more , it guarantees a high multiplicity bh decay .
the bh decays more or less isotropically and each decay particle has an average energy of a few hundred gev .
therefore , if the bh is at rest , the event is very much like a spherical event with many particles of hundreds of gev pointing back to the interaction point ( very much like a fireball ) . on the other hand ,
if the bh is produced in association with other sm particles ( as in a @xmath174 subprocess ) , the bh decay will be a boosted spherical event on one side ( a boosted fireball ) , the transverse momentum of which is balanced by a few particles on the other side @xcite .
such spectacular events should have a negligible background .
highly excited long strings emit massless quanta with a thermal spectrum at the _
hagedorn _ temperature .
( the hagedorn temperature of an excited string matches the hawking temperature of a bh at the corresponding point @xmath175 . ) at @xmath176 , the wavelength @xmath166 corresponding to the thermal spectrum at the hagedorn temperature is larger than @xmath177 .
this argument is very similar to that of the bh , and so the string ball radiates like a point source and emits in @xmath81-waves equally into brane and bulk modes . with many more particles ( sm particles ) on the brane
than in the bulk , the sb radiates mainly into the sm particles .
when @xmath178 goes below @xmath2 , the sb has the tendency to puff up to a _ random - walk size _ as large as the @xmath166 of the emissions @xcite .
therefore , it will see more of the higher angular momentum states available in the extra dimensions .
thus , it decays more into the bulk modes , but it is only temporary .
when the sb decays further , it shrinks back to the string size and emits as a point source again @xcite .
most of the time the sb decays into sm particles . on average
, a sb decays into invisible quanta somewhat more often than a bh does .
high multiplicity decay of the bh should also apply to the sb , at least when the mass of the sb is close to the correspondence point @xcite .
naively , we expect that if the sb mass decreases , the multiplicity will decrease .
thus , the signature of the sb is very similar to the bh , except that it may have lower multiplicity .
the decay of @xmath0-branes is not well understood , to some extent we do not even know whether it decays or is stable . nevertheless , if it decays one possibility is the decay into lower - dimensional branes , thus leading to a cascade of branes . therefore , they eventually decay to a number of @xmath12-branes , i.e. , bh - like objects .
this is complicated by the fact that when the @xmath0-branes decay , their masses might not be high enough to become bh s .
therefore , the final @xmath12-branes might be some excited string states or string balls . whether the zero brane is stable or not depends on models .
another possibility is decay into brane and bulk particles , thus experimentally the decay can be observed . or it can be a combination of cascade into lower - dimensional branes and direct decays . since the size @xmath106 is much smaller than the size of the large extra dimensions , we expect @xmath0-branes to decay mainly into brane particles .
however , the above is quite speculative .
in this work , we have calculated and compared the production cross sections for black holes , string balls , and @xmath0-branes at hadronic supercolliders ( lhc and vlhc ) .
provided that the fundamental planck scale is of order of 1 to a few tev , large numbers of bh , sb , and @xmath0-brane events should be observed at the lhc . at the vlhc ( @xmath179 tev ) ,
the events rates are enormous .
we have also given rough estimates for the sensitivity reaches on the fundamental planck scale @xmath4 at various @xmath125 , based on the number of raw events .
the sensitivity of bh and @xmath0-brane production is roughly 2 tev at the lhc , @xmath180 tev for @xmath152 tev , @xmath181 tev for @xmath155 tev , @xmath182 tev for @xmath158 tev , and @xmath183 tev for @xmath161 tev .
finally , we offer a few comments as follows . 1 . the production cross sections for bh estimated in this work are significantly smaller than others in the literature , because we have imposed a stringent entropy @xmath7 requirement on the bh .
such a requirement is necessary to make sure the object is a bh . had
this requirement relaxed , the cross section would have increased substantially . for the purpose of comparing with others results , we also show the cross sections for smaller values of @xmath184 in table [ table - y ] .
the cross sections listed for @xmath185 should be interpreted with care , because the smaller the ratio @xmath186 the stronger the string effect is and the classical description for bh may not be valid . 2 .
it was pointed out in ref .
@xcite that a bh with an angular momentum @xmath187 is likely to be formed in particle collisions when the incoming partons are collided at an impact parameter . in such a case , the radius of the bh decreases and thus the naive cross - section formula @xmath188 implies a smaller cross section for each angular momentum @xmath187
. the higher the angular momentum , the larger is the suppressions .
nevertheless , when all @xmath187 ( including @xmath189 ) are summed , the total cross section gives a factor of @xmath190 enhancement to the case of nonspinning bh .
@xmath0-brane production is negligible if @xmath191 , because of the large volume factor suppression .
but when @xmath16 ( the @xmath0-brane wraps entirely on the small extra dimensions of the size of the planck length ) , the production cross section is sizable .
moreover , the cross section is a few times larger than the bh production for the case of @xmath192 , where @xmath193 .
the production cross section for sb s is enormous because it does not suffer from a mass threshold as large as for the bh .
the minimum mass requirement is between @xmath1 and @xmath80 .
we typically choose @xmath119 as the starting point for the sb .
such a large event rate makes the tests for string ball properties and bh correspondence principle possible . as pointed out in ref .
@xcite , since only a very small fraction of the decay products of a bh has missing energies , the mass of the bh can be determined .
moreover , the energy spectrum of the decay products can be measured and fitted to the black - body radiation temperature .
thus , the hawking radiation relationship between the mass and temperature of a bh can be tested . here ,
similar to the bh , both the mass and the temperature of the sb can be determined by measuring the spectrum of the decay products .
thus , the relationship between the mass of the sb and the hagedorn temperature can be tested .
we have emphasized the importance and the advantages of using the @xmath120 subprocess for production of bh s , sb s , and @xmath0-branes , which allows a substantial transverse momentum kick to the object , and at the same time produces an energetic high @xmath52 parton , which provides a critical tag to the event . 6 . at the lhc and vlhc , multiparton collisions and overlapping events
may be likely to happen .
a careful discrimination is therefore necessary , especially in the case in which the bh is produced at rest or is moving along the beam - pipe ( i.e. , in @xmath49 subprocess ) .
the @xmath120 subprocess affords an easier signature experimentally .
the high @xmath52 parton emerging as a jet , a lepton , jets , or leptons provides an easy tag .
7 . in this study
, we do not consider the difference in the decays of bh , sb , and @xmath0-brane .
if we could distinguish the decay signatures of the bh and sb , we might be able to test the bh correspondence principle at the transition point .
we can also test the decays of @xmath0-branes in more detail .
there just appears a short review article @xcite on bh production at hadronic colliders and by ultrahigh energy cosmic neutrinos .
i would like to thank jeonghyeon song , kang young lee , yong yeon keum , and eung jin chun for an invitation to the korea institute for advanced study , where part of this work was done , and for their hospitality .
also thanks to seong chan park for an interesting discussion .
this research was supported in part by the national center for theoretical science under a grant from the national science council of taiwan r.o.c . n. arkani - hamed , s. dimopoulos , and g. dvali , phys .
lett * b429 * , 263 ( 1998 ) ; i. antoniadis , n. arkani - hamed , s. dimopoulos , and g. dvali , phys . lett . *
b436 * , 257 ( 1998 ) ; n. arkani - hamed , s. dimopoulos , and g. dvali , phys . rev . * d59 * , 086004 ( 1999 ) .
giddings , hep - ph/0110127 .
r. casadio and b. harms , hep - th/0110255 .
s. hofmann et al . , hep - ph/0111052 ; s. park and h. song , hep - ph/0111069 ; g. landsberg , phys .
lett . * 88 * , 181801 ( 2002 ) .
g. giudice , r. rattazzi and j. wells , nucl . phys . *
b630 * , 293 ( 2002 ) .
m. bleicher et al .
, hep - ph/0112186 ; s. solodukhin , phys . lett . * b533 * , 153 ( 2002 ) . t. rizzo , jhep * 0202 * , 011 ( 2002 ) .
l. anchordoqui , h. goldberg , and a. shapere , hep - ph/0204228 .
p. jain , d. mckay , s. panda , and j. ralston , phys .
b484 * , 267 ( 2000 ) ; j. feng and a. shapere , phys .
lett . * 88 * , 021303 ( 2002 ) ; l. anchordoqui , j. feng , h. goldberg , and a. shapere , phys . rev . *
d65 * , 124027 ( 2002 ) .
l. anchordoqui and h. goldberg , phys . rev .
* d65 * , 047502 ( 2002 ) .
r. emparan , m. masip , and r. rattazzi , phys . rev . *
d65 * , 064023 ( 2002 ) .
a. ringwald and h. tu , phys .
* b525 * , 135 ( 2002 ) . y. uehara , prog .
. phys . * 107 * , 621 ( 2002 ) .
m. kowalski , a. ringwald , and h. tu , phys . lett .
* b529 * , 1 ( 2002 ) . j. alvarez - muniz , j. feng , f. halzen , t. han , and d. hooper , phys . rev . * d65 * , 124015 ( 2002 ) .
m. voloshin , phys . lett . *
b518 * , 137 ( 2001 ) ; phys . lett . *
b524 * , 376 ( 2002 ) .
r. casadio and b. harms , phys . lett .
* b487 * , 209 ( 2000 ) ; phys . rev . *
d64 * , 024016 ( 2001 ) .
p. kanti and j. march - rusell , hep - ph/0203223 .
p. kanti and k. tamvakis , phys
* d65 * , 084010 ( 2002 ) .
h. kim , s. moon , and j. yee , jhep * 0202 * , 046 ( 2002 ) .
d. eardley and s. giddings , gr - qc/020134 .
a. jevicki and j. thaler , hep - th/0203172 .
s. hsu , hep - ph/0203154 .
s. bilke , e. lipartia , and m. maul , hep - ph/0204040 .
v. frolov and d. stojkovic , e - print archive : hep - th/0206046 .
s. dimopoulos and r. emparan , phys . lett . *
b526 * , 393 ( 2002 ) .
ahn , m. cavaglia , and a. olinto , hep - th/0201042 .
p. jain , s. kar , s. panda , and j. ralston , hep - ph/0201232 .
l. anchordoqui , j. feng , and h. goldberg , phys .
b535 * , 302 ( 2002 ) .
k. oda and n. okada , hep - ph/0111298 .
j. friess , t. han , and d. hooper , hep - ph/0204112 .
g. shiu and s. tye , phys . rev . *
d58 * , 106007 ( 1998 ) .
m4 working group / hadron colliders `` , plenary talk given by m. syphers at the snowmass 2001 , available online at http://www.vlhc.org/m4finalplenary.pdf .
snowmass 2001 ' ' the future of particle physics " meeting , snowmass , summer 2001 .
the subprocess cross section @xmath195(sb / bh ) for string ball or black hole vs the mass of sb or bh .
here we have used a string scale @xmath196 tev , and we require the sb - bh correspondence point at @xmath197 , where @xmath4 is related to @xmath1 by @xmath198 . ,
width=480 ] total production cross section @xmath199 including @xmath111 and @xmath136 processes for black hole ( bh ) , string ball ( sb ) , and @xmath0-brane ( @xmath0b ) at @xmath145 collisions vs @xmath125 for @xmath121 and @xmath148 .
here we have used a fundamental scale @xmath55 tev .
the minimum mass on the bh and @xmath0-brane is @xmath200 , while that on sb is @xmath201 . @xmath202 and @xmath203 tev for @xmath121 and @xmath148 , respectively , in our scheme .
, width=480 ] differential cross section @xmath204 vs the mass @xmath205 of black hole ( bh ) , string ball ( sb ) , or @xmath0-brane ( @xmath0b ) at the lhc .
here we have used a fundamental scale @xmath55 tev and @xmath56 .
the minimum mass on the bh and @xmath0-brane is @xmath206 , while that on sb is @xmath201 .
@xmath207 tev for @xmath56 . , width=480 ]
differential cross section @xmath208 vs the transverse momentum @xmath52 of black hole ( bh ) , string ball ( sb ) , or @xmath0-brane ( @xmath0b ) at the lhc .
here we have used a fundamental scale @xmath55 tev and @xmath56 .
the minimum mass on the bh and @xmath0-brane is @xmath206 , while that on sb is @xmath201 .
@xmath207 tev for @xmath56 .
, width=480 ] total production cross section @xmath199 including @xmath111 and @xmath136 processes for black hole ( bh ) , string ball ( sb ) , and @xmath0-brane ( @xmath0b ) at @xmath145 collisions vs @xmath125 from @xmath209 tev for @xmath121 and @xmath148 .
here we have used a fundamental scale @xmath210 tev .
the minimum mass on the bh and @xmath0-brane is @xmath200 , while that on sb is @xmath201 . @xmath211 and @xmath212 tev for @xmath121 and @xmath148 , respectively , in our scheme . ,
width=480 ] .[pbrane ] the ratio @xmath213 of eq .
( [ r ] ) for various @xmath21 and @xmath0 with @xmath214 .
we have used @xmath55 tev and @xmath215 .
we have assumed that the @xmath0-brane wraps entirely on small extra dimensions , i.e. , @xmath16 . in order to obtain the largest ratio @xmath216 we have chosen @xmath217 . [
cols="^,^,^,^,^,^,^",options="header " , ] | in models of large extra dimensions , the string and planck scales become accessible at future colliders .
when the energy scale is above the string scale or planck scale a number of interesting phenomena occur , namely , production of stringy states , @xmath0-branes , string balls , black hole , etc . in this work
, we systematically study the production cross sections of black holes , string balls , and @xmath0-branes at hadronic supercolliders .
we also discuss their signatures . at the energy scale between the string scale @xmath1 and @xmath2 , where @xmath3 is the string coupling ,
the production is dominated by string balls , while beyond @xmath2 it is dominated by black holes .
the production of a @xmath0-brane is only comparable to black holes when the @xmath0-brane wraps entirely on small extra dimensions .
rough estimates on the sensitivity reaches on the fundamental planck scale @xmath4 are also obtained , based on the number of raw events . |
cosmology ( i.e. the modern theory of universe dynamics ) has became in the last two decades an attractive field of human knowledge including also an intense media campaign .
this was possible , among other reasons , also because in this time period several space missions performed cosmological measurements -like cobe , wmap , plank , bicep2- thus transforming cosmology from a pure theoretical field also into an experimental one . thus teaching cosmology , even at undergraduate level , in physics faculties comes to be a compulsory topic .
cosmology is based on two major pillars @xcite,@xcite : astrophysics as a phenomenological tool and general relativity ( gr ) as the main theory , thus making it , unfortunately , difficult to teach to undergrad students . the mathematical structure or gr is based on differential geometry @xcite , @xcite and learning it means that the student must first get familiar with the main instruments of differential geometry ( such as tensor calculus on curved manifolds , riemannian curvature , metric connection , covariant derivative , etc ) , which often means cumbersome and lengthy hand calculations . to illustrate these facts let us remind that gr is based on field equations known as einstein equations , namely @xmath1 where @xmath2 is the ricci tensor , @xmath3 the metric tensor and @xmath4 represents the stress - energy tensor .
we denoted by @xmath5 the cosmological constant and @xmath6 .
the components of the ricci tensor are given by @xcite,@xcite @xmath7 where @xmath8 are the so called chrisstoffell symbols wich in riemannian geometry @xcite describe the structure of the curved space - time underlying gr and the associated metric tensor is compatible with the connection described by the above chrisstoffel symbols , namely @xmath9 all these make even the most determined students to lose interest in studying cosmology ( and gr too ) .
dozens of pages with hundreds of terms containing partial differentials to be hand processed could scare any student ( and not only ! ) . as today
students are by passing day more skilled in computer manipulation and with a more and more advanced practice in programming , the modern teaching of gr and cosmology should use intensively computer facilities for algebraic programming , tensor manipulation and of course numerical and graphical facilities @xcite , @xcite , @xcite .
the use of computer algebra was in the view of physicists even since the beginning of computer science both for teaching and research purposes .
computer algebra ( or algebraic programming codes ) evolved from early days of reduce package ( see for example @xcite,@xcite , @xcite ) till recent developments using integrated platforms as maple and mathematica in different fields of physics , not only in general relativity ( see for example @xcite , @xcite , @xcite and more recently @xcite , @xcite ) . some years ago
@xcite we published our experience in this using the reduce platform .
unfortunately , in the last years reduce lost the market in favour of more intergraded and visual platforms such as maple @xcite and mathematica , thus we adapted our experience and program packages to maple and we ve made use of the free package gttensorii @xcite adapted for doing gr .
the aim of the present article is to report our new experience in this direction .
the article is organised as follows .
the next section introduces the main features of grtensorii in doing tensorial symbolic computation in gr and riemannian geometry .
section no .
3 describes the way we can obtain an exact solution of einstein equations .
we used again , as in the main classical text on gr , the schwarzschild solution .
this is the most famous solution used today intensively in describing the motion in the solar system and for studying black - holes physics @xcite .
the last section is dedicated to describe how one can use maple and grtensorii for cosmology ( and teaching it ) . in both these two above sections we gave the main maple commands which can be put together to have short programs to be used during the computer lab hours and even during the lectures .
the article ends with a short section where we present the main conclusions and some ideas for future developments .
grtensorii is a computer algebra package built within the maple platform as a special set of libraries @xcite .
it is a free distributed package ( see @xcite ) and it is adapted for dealing with computer algebra manipulation in general relativity .
thus it is designed for dealing with tensors and other geometric objects specific to riemannian geometry ( a metric that is compatible with the connection , symmetric connection - torsion free manifolds ) . in what follows , we will present some of the main features offered by grtensorii .
the library is based on a series of special commands all starting with @xmath10 ( for example @xmath11 , etc ) for dealing with a series of ( pre)defined geometric objects such as the metric tensor , ricci tensor and scalar , einstein tensor , chrisstoffell symboles , etc . to start the grtensorii package one must type in a ( new ) maple session the following commands .... > restart ; > grtw ( ) ; .... the restart command causes the maple kernel to clear its internal memory so that maple acts ( almost ) as if just started .
the second command initializes the grtensorii package and gives some information about the current version .
the grtensorii library allows us to build our own space - time manifold .
the most easiest way to do this is by creating a metric tensor @xmath12 with the help of @xmath13 command .
the result of operating this command will be a special ascii file containing the main information about the metric and stored in a folder called `` metrics '' within the grtensorii library .
the folder `` metrics '' contains also a collection of predefined metric files distributed with the package ( for egz .
schw.mpl , schmidt.mpl , vdust.mpl , etc ) .
another way in which one can specify a space - time manifold is by loading a predefined metric from the `` metrics '' folder .
this can be done with the help of two commands .... > qload(metricname ) ; > grload(metricname , meticfile ) ; .... the geometry built using @xmath14 , @xmath15 and @xmath16 commands will fix the background on which all the later operations and calculations will be performed .
one of the main advantages of grtensorii library is that it allows us to do complicated operations on tensorial objects , regardless of how many indices those objects poses .
the main command that permit us to do those calculations is @xmath17 that calculates the components of tensors ; for example @xmath18 ask the program to calculate the covariant components of the ricci tensor @xmath19 and the components of the standard curvature riemann tensor @xmath20 . the command @xmath21 can be used to display the components of grtensorii objects which have been previously calculated for a particular space - time . before displaying the calculated components of an object it is indicated to use the command @xmath22 in order to simplify them .
the @xmath23 calculates all the components of a given object , so if one wants to calculate only a specific component then it can use the command @xmath24 .
for example : @xmath25)$ ] . besides the predefined objects that exist in grtensorii we can also define new objects ( scalars , vectors , tensors ) with the help of the command @xmath26 . for example @xmath27 defines a contravariant two index tensor , @xmath28 , which it is explicitly assigned to an expression involving a number of previously defined ( or predefined ) tensors .
the syntax in @xmath29 command follows naturally the usual tensorial operations which defines the new objects . another important command of grtensorii is @xmath30 which allows us to extract a certain component of a tensorial object .
the extracted component can be used as a standard maple object for later processing ( symbolically , graphically and numerically ) .
although grtensorii was designed initially for riemannian differential geometry it can be easily extended to other types of geometries , such as ones with torsion or higher order alternative theories of gravity @xcite , @xcite .
we end this section by making the important observation that using grtensorii does not impose any restriction in using all the numerical , graphical and symbolic computation facilities of maple ( as it happens with other packages even for maple ) .
thus we can combine al these facilities for an efficient use of the maple platform .
general relativity and its applications ( such as cosmology ) are based on einstein equations ( [ ee ] ) as main field equations .
they have many exact solutions , although these second order nonlinear differential equations have no unique analytical general solution . the most famous exact solution of einstein equations is the schwarzschild solution @xcite , @xcite describing the gravitational field around a pointlike mass m ( or outside a sphere of mass m ) .
this solution is used today for describing the black - hole dynamics and was used in the first attempts in applying gr to the motion of planets and planetoids in our solar system .
it is obvious that from a pedagogical point of view finding an exact solution of einstein equations could be a good introductory lesson in applications of gr .
next we will derive this solution following the natural steps : - identifying the symmetries of the system for which we build the solution ; - building a metric tensor compatible with the above symmetry ; - building the shape of the stress - energy tensor components ( if any exists ) ; - calculating the ricci tensor and the components of einstein equations ; - solving the above equations after a close inspection of them . these above steps could be done manually and usually it takes several hours of hard calculations ( even straightforward and even for an experienced person ) . our advise for anyone who wants to teach gr and/or cosmology is to do this traditional step with the students .
it will be a good lesson and a motivation to proceed in using algebraic computing facilities ( here maple+grtensorii ) .
thus the above steps are clearly transposable in computer commands in maple+ grtensorii . for the schwarzschild solution
the symmetry is clearly spherical and static ( no time dependence including the time inversion , namely @xmath31 ) .
thus we will use a spherical symmetric metric tensor as @xcite : @xmath32 in spherical coordinates @xmath33 , where @xmath34 and @xmath35 are the two unknown functions of the radial coordinate @xmath36 to be found at the end .
thus the student already in front of a computer or station having started a maple session will be guided to compose the next sequence of commands .... > restart ; grtw ( ) ; >
makeg(sferic ) ; > ...... > grdisplay(metric ) ; grdisplay(ds ) ; .... where after the two commands for starting the grtensortii the command @xmath37 will create the ascii file @xmath38 containing the information on the metric we will built .
the series of dots above represent those steps where the user has to answer with the type of the metric , symmetry and of course its components one by one .
the last two lines given above are calculating the metric and displays its shape in the form of a matrix .
after this we can continue to do some calculations or to close the session .
the metric we produce can be loaded anytime later in another sessions .
the next step of our demonstrative program will be to point out and calculate the einstein equations , but not before introducing the stress - energy components .
it is obvious that in this case the strass - energy components are cancelled as we calculate the gravity field outside the source ( the pointlike mass or a sphere ) .
thus in this case we will solve the so called vacuum einstein equations i.e. @xmath39 @xcite .
in this view we will built a sequence of maple commands ( for a new session ) : .... > restart ; grtw ( ) ; > qload(sferic ) ; grcalc(r(dn , dn ) ) ; > gralter(r(dn , dn),simplify ) ; grdisplay(r(dn , dn ) ) ; .... where after loading the metric tensor with the command @xmath40 we calculate the ricci tensor components , simplify and display them as a 4x4 matrix ( using the last @xmath41 command ) .
it is a good and interesting experience , before proceeding with the solving of einstein eqs . to insert in the above lines the next ones ....
> grcalc(chr(up , dn , dn ) ) ; grdisplay(chr(up , dn , dn ) ) ; .... immediately after loading the metric tensor with @xmath42 which very fast calculates and displays the 40 components of the chrisstoffel symbols using relation ( [ chris ] ) .
of course if we want maximum effect of these , before calculating with maple and grtensorii we advise the teacher to calculate , by hand , together with the students at least 3 of 4 of these components .
this will take some time of hard calculations and many mistakes when done for the first time . to continue
it is now much more simple to extract the components of the ricci tensor one by one as maple objects in order to process them to solve the obtained equations .
it will be a sequence of @xmath43 commands , namely : .... > ec0:=grcomponent(r(dn , dn),[t , t ] ) ; > ec1:=grcomponent(r(dn , dn),[r , r ] ) ; > ec2:=grcomponent(r(dn , dn),[theta , theta ] ) ; >
ec3:=grcomponent(r(dn , dn),[phi , phi ] ) ; .... obtaining the four einstein equations of the problem .
the rest of the ricci tensor components are zero .
a simple inspection of the above obtained four equations reveals that only two of them are independent .
also we can eliminate the second order derivative of the @xmath34 function between @xmath44 and @xmath45 .
these can be checked by using the next command lines .... > expand(simplify(ecu2-ecu3/sin(theta)^2 ) ) ; > ecu0;ecu1;ecu2 ; > l2r:=solve(subs(diff(lambda(r),r ,
r)=l2r , ecu0),l2r ) ; > expand(simplify(subs(diff(lambda(r),r , r)=l2r , ecu0 ) ) ) ; > ecu11:=expand(simplify(subs(diff(lambda(r),r , r)=l2r , ecu1 ) ) ) ; .... the first of the above commands checks the equality between @xmath46 and @xmath47 , which simply gives a `` 0 '' ( zero ) and the next ones simply display the remaining three equations . the command that follows extracts the second order derivative of @xmath34 from @xmath48 ( substituting it with an intermediate constant @xmath49 ) and the next one substitutes the result in @xmath50 and we obtained a new equation @xmath51 .
thus we have now only two equations , @xmath51 and @xmath46 , namely @xmath52 @xmath53e^{-2\mu(r)}=0\ ] ] these two differential equations need to be solved next , in order to obtain the solution .
of course we can now follow a classical strategy solving them manually as is done in any textbook ( see @xcite for example ) .
but it is also possible to continue with maple , using the @xmath54 command for solving differential equations ( including systems of differential equations ) .
thus we write the next command .... > dsolve({ecu11,ecu2},{mu(r),lambda(r ) } ) ; .... which gives us the function
@xmath35 as @xmath55 in particular a close inspection of both equations @xmath56 reveals that @xmath51 is simply a relation between the derivatives of the two functions , namely @xmath57 this shows that we will need only one integration constant and we can solve the equations only for one function as done above .
one can write the constant @xmath58 as @xmath59 where the new introduced constant @xmath60 will be determined later . with these we can rewrite the @xmath46 and applying again the @xmath54 command on it we obtain in the schwarzschild ( type ) solution as @xmath61\ ] ]
the constant @xmath60 is known under the name of schwarzschild radius and can be determined using the newtonian limit of the field equations as it is done in any textbook ( see @xcite for example ) .
the precise value of it is @xmath62 but this has nothing to do with algebraic computing .
modern cosmology is based on general relativity and einstein equations , from which we can derive the so called friedman equations .
the latter ones form the core of all cosmological models .
the most used metric for describing the dynamics of the universe in a cosmological model is the friedman - robertson - walker metric ( frw ) , which in spherical coordinates has the following line element @xcite @xmath63\ ] ] where @xmath64 is the curvature constant and we are using the @xmath65 signature for the metric . usually , this @xmath64 constant is taken to be @xmath66 ( for closed universes ) , @xmath67 ( for open universe ) and @xmath68 for a flat one . in ( [ cm1 ] )
we denoted by @xmath69 the scale factor , that in the end will be the only unknown function of a cosmological model .
this scale factor is directly related to the evolution of the universe . for the frw metric ( obtained form
the cosmological principle i.e. the universe is spatially homogenous and isotropic ) the scale factor is a function only on time . by introducing frw metric ( [ cm1 ] ) into the einstein equations ( [ ee ] ) and assuming that the energy - momentum tensor is of a prefect fluid form @xcite @xmath70 one arrives to the friedman - lemaitre equations @xmath71 if the cosmological constant @xmath5 is set to zero in eqs.([rm4 ] )
then the equations are called simply the friedman equations . in eq .
( [ rm3 ] ) @xmath72 represnets the 4-velocity of the cosmological fluid , while @xmath73 and @xmath74 stands for the pressure , respectively the mass density of the fluid . in the same manner as done in @xcite we can compose a sequence of grtensorii commands for obtaining the friedman equations ( [ rm4 ] ) .
a student can write the program on a computer in less than an hour to the a job that if it is done by hand calculations it will take several good hours to a very good student
. the basic lines of the grtensorii program are as follows : .... > restart ; > grtw ( ) ; qload(metrica_frw ) ; > grdef(`u{^a}:=[1,0,0,0 ] ` ) ; >
grdef(`t{^a ^b}:=(rho(t)+p(t)/c^2)*u{^a}*u{^b}-p(t)*g{^a ^b } ` ) ; > ... > ec0:=r1_0-t1_0 ; > ec1:=r1_1-t1_1 ; > ec1:=subs(diff(a(t),t , t)=-(1/6)*k*c^4*rho(t)*a(t)- ( 1/2)*k*c^2*p(t)*a(t)+(1/3)*lambda*c^2*a(t),ec1 ) : ec0 ; ec1 ; .... the first line of commands starts the grtensorii and loads the frw metric ( [ cm1 ] ) . in the next lines we define the 4-velocity and the stress - energy tensor ( [ rm3 ] ) of the cosmological fluid . now follows a series of more technical commands which can be found in the supplementary web material @xcite , commands that allows us to calculate and write the final form of friedman equations ( [ rm4 ] ) .
friedman equations ( [ rm4 ] ) are in fact a system of two differential equations with three unknowns : @xmath69 , @xmath73 and @xmath74 .
thus one needs to find a third equation in order to completely solve the problem . in standard cosmology
we use as a third relation the equation of state @xmath75 where @xmath76 is a constant ( @xmath77 for pressureless dust , @xmath78 for radiation and @xmath79 for vacuum ) .
let us further introduce the dimensionless quantities ( see for example @xcite ) , usually called density parameters , which are defined by @xmath80 where @xmath81 is known as the hubble parameter and @xmath82 stands for matter , radiation and the cosmological constant @xmath5 .
besides these three quantities one can also define a curvature density parameter @xmath83 rewriting the second equation of ( [ rm4 ] ) in terms of the new defined density parameters we arrive at a very simple expression @xmath84 taking all the above into account and introducing a normalised scale factor @xmath85 ( with @xmath86 representing the value of the scale factor at the present epoch ) we can finally write an equation for the evolution of the scale factor , namely @xmath87 where @xmath88 are the values of the densities measured at the present epoch .
below we give the main grtensorii command lines with the help of which we arrive at equation ( [ rm7 ] ) .
the complete sequence of commands can be found in the supplementary web material @xcite . ....
> ec1:=subs(diff(a(t),t)=h(t)*a(t),ec1 ) ; > rho(t):=rho_m0*(a_0/a(t))^3+rho_r0*(a_0/a(t))^4 ; >
ec1:=subs(omega_k0=1-omega_m0-omega_r0-omega_lambda0,ec1 ) ; .... we can now use the other numerical and computational facilities of maple in order to numerically solve equation ( [ rm7 ] ) and express the results as plots ( see fig .
1 ) of the scale factor as a function of cosmic time . for
that we use the following code ( see also the supplementary web material @xcite ) : .... >
ecu:=diff(a(t ) , t)-sqrt(omega_m0/a(t)+omega_r0/a(t)^2 + omega_lambda0*a(t)^2 + 1-omega_m0-omega_r0-omega_lambda0 ) ; > ecu_a:=subs(omega_m0 = 0.3,omega_lambda0=0.7,omega_r0=0,ecu ) ; >
sys1:={ecu_a , a(0)=1 } : > f1:=dsolve(sys1,numeric ) : >
odeplot(f1,t=-2 .. 2,axes = boxed , numpoints=1000,color = black ) ; .... 0.47 0.47
the article describes a way in which some simple computer programs in maple and grtensorii can be used in teaching gr and cosmology .
it is obvious that the speed of learning main concepts in gr ( and subsequently differential geometry ) can be successfully enhanced , avoiding large hand computation steps and a lot of natural mistakes . on the other hand
it is clear that the lectures will need to take place in a computer lab , which can be another way to increase the attractiveness of gr and cosmology .
we illustrated our experience with short and simple commands without sophisticated tricks normally a professional in the field is using ( for example building procedures and libraries via the symbolic computation facilities of maple ) .
the small and short programs we described here can be also used as a strong basis for further developments in view of more sophisticated and advanced examples .
for instance one can develop the above procedures for cosmology in generalised theories of gravity , like those with higher order lagrangians ( as we done in @xcite ) .
this work was supported by a grant of the romanian national authority for scientific research , programme for research - space technology and advanced research - star , project nr .
72/29.11.2013 between romanian space agency and west university of timisoara .
c.a.sporea was supported by the strategic grant posdru/159/1.5/s/137750 , project `` doctoral and postdoctoral programs support for increased competitiveness in exact sciences research '' cofinanced by the european social found within the sectorial operational program human resources development 2007 2013 .
trandafir et al . , `` elementary tight - binding method for simple electronic structure calculations - an educational approach to modeling conjugated dyes for dye - sensitized solar cells '' , rom .
66 , 574 ( 2014 ) .
vulcanov , g.s .
djordjevic , c.a .
sporea , `` rem - the shape of potentials for f(r ) theories in cosmology and tachyons '' , in cosmology and particle physics beyond standard models .
ten years of the seenet - mtp network , edited by luis alvarez - gaume , goran s. djordjevic , dejan stojkovc , cern - proceedings-2014 - 001 , pp .
165 - 169 . | in this article we propose some maple procedures , for teaching purposes , to study the basics of general relativity ( gr ) and cosmology . after presenting some features of grtensorii , a package specially built to deal with gr , we give two examples of how one can use these procedures . in the first example we build the schwarzschild solution of einstein equations , while in the second one we study some simple cosmological models .
+ * keywords : * general relativity @xmath0 grtensorii @xmath0 cosmology * pacs ( 2010 ) : * 98.80.-k @xmath0 04.30.-w @xmath0 01.40.ha + |
by a _ polyhedron _ , denoted by @xmath5 , we mean the boundary of a compact and convex polyhedron in @xmath6 .
@xmath5 is naturally endowed with its _ intrinsic metric _ :
the distance between two points is the length of the shortest curve joining them . in this paper
, we shall never consider the extrinsic distance .
a _ segment _ is by definition a shortest path on the polyhedron between its endpoints . in general , it is not a line segment of @xmath6 , but becomes so if one unfolds the faces it crosses onto a same plane
. an _ antipode _ of @xmath2 is a farthest point from @xmath2 ; the set of antipodes of @xmath2 is denoted by @xmath3 .
it is well - known that the mapping @xmath1 is upper semicontinuous .
when the context makes clear that @xmath3 is a singleton , we shall not distinguish between this singleton and its only element .
the study of antipodes on convex surfaces began with several questions of h. steinhaus , reported in @xcite , most of them answered by tudor zamfirescu , see @xcite , @xcite , @xcite , @xcite .
however , one of those questions had remained open a little longer : does the fact that the antipodal map of a convex surface is a single - valued involution imply that the surface is a round sphere ? as we shall see , the answer is negative . by definition , such a surface will be called a _
steinhaus surface_. the first family of steinhaus surfaces was discovered by c. vlcu @xcite .
it consists of centrally symmetric surfaces of revolution , and includes the ellipsoids having two axes equal , and the third shorter than the two equal ones .
( note that if the third axis is longer than the two equal ones , the surface is no longer steinhaus @xcite . ) other examples were discovered afterward : cylinders of small height @xcite , and the boundaries of intersections of two solid balls , provided that the part of the surface of the smaller ball included in the bigger one does not exceed a hemisphere .
this last example , as well as the mentioned family , were generalized to hypersurfaces in @xmath7 @xcite .
note that steinhaus conditions are related to another one .
define the _ radius _ at a point @xmath2 as the distance between @xmath2 and its antipodes .
it is known that , if some surface has a constant radius map , then it is a steinhaus surface @xcite .
moreover , all examples of steinhaus surfaces hitherto discovered also satisfy the constant radius condition .
so it is still open whether the two conditions are equivalent .
one can also notice that all known examples are surfaces of revolution , and in particular , are not polyhedral .
the first attempt to find a polyhedral example was the investigation of the regular tetrahedron .
an explicit computation of the antipodal map proved that it is not a steinhaus surface @xcite .
then we proved that no tetrahedra can be steinhaus @xcite .
a few years later , we proved that no polyhedron can satisfy the radius condition , and that no centrally symmetric polyhedron can be steinhaus @xcite . since then , the problem of the ( non)existence of steinhaus polyhedra was very natural .
the aim of this paper to prove they do not exist . in order to make this article
self - contained , we give in section [ s2 ] some preliminary , concerning the antipodes on a convex polyhedron . in section [ s3 ] ,
we prove the result .
then , in a last section , we discuss a few open questions related to this topic .
the explicit computation of the antipodal map in the case of the regular tetrahedron @xcite was generalized as much as possible to arbitrary convex polyhedra @xcite,@xcite .
we proved there the following theorem .
[ t0]any convex polyhedron @xmath5 can be written as a disjoint union@xmath8 with the following properties . 1 . the sets @xmath9 are open , and the restricted maps @xmath10 are singled - valued .
2 . the set @xmath11 is a finite union of algebraic arcs , of degree at most @xmath12 .
it includes the edges of @xmath5 .
3 . the map @xmath10 is a constant map if and only if its image is a singleton containing a vertex .
if @xmath13 , then there is exactly one segment between @xmath14 and any point of @xmath9 ; otherwise , there are exactly three segments between @xmath15 and @xmath3 . from now on , the sets
@xmath9 defined in the above theorem will be referred to as _
zones_. including a priori the edges in @xmath11 is only a trick to simplify the proof and ensure that any zone is ( isometric to ) a plane domain . in general
, nothing special happens while crossing an edge : if @xmath16 and @xmath17 are two zones separated by an edge and one rotates the face of @xmath16 onto the plane of @xmath17 , then @xmath18 and @xmath19 become rational prolongations of each other .
more generally , the notion of edges belongs to the extrinsic geometry and plays no role in a purely intrinsic problem .
the proof of theorem [ t0 ] is given in details in @xcite and @xcite , we present next only its rough idea .
the fourth point of the theorem has been stated for aim of completeness only .
it will not be used here , so we shall omit its proof which involves extensive computation .
a first remark is that a segment can not pass through any vertex , see for instance @xcite .
secondly , note that a point @xmath2 on the polyhedron is joined to any of its antipodes @xmath20 by at least three segments , provided that @xmath20 is not a vertex .
suppose on the contrary that there are only two segments @xmath21 and @xmath22 between @xmath2 and @xmath20 .
they would determine two sectors at point @xmath20 , one of them having measure at least @xmath23 ( if there is only one segment , there is only one sector , measuring @xmath24 ) .
consider a point @xmath25 on the bisector of this sector , tending to @xmath20 .
a segment @xmath26 between @xmath2 and @xmath25 should tend to either @xmath21 or @xmath22 , say @xmath21 .
for @xmath25 close enough to @xmath20 , the triangle composed of @xmath21 , @xmath26 , and the only segment between @xmath25 and @xmath20 contains no vertex , and so , is a ( folded ) euclidean triangle .
moreover the angle at @xmath20 is obtuse or right ; it follows that the length of @xmath26 ( which is also @xmath27 ) is greater than the length of @xmath21 ( which is supposed to be the radius at @xmath2 ) , and we get a contradiction .
hence there exist three segments @xmath28 , @xmath29 , @xmath21 between @xmath2 and @xmath20 .
let @xmath30 be the face of @xmath2 and @xmath31 be the face of @xmath20 ; if one unfolds the union of the faces crossed by @xmath32 ( @xmath33 , @xmath34 , @xmath35 ) onto the plane of @xmath31 , one obtains three images of the face @xmath30 , say @xmath36 , @xmath37 and @xmath38 .
one passes from @xmath36 to @xmath39 ( @xmath40 ) by a planar affine direct isometry @xmath41 .
since the segments @xmath32 have the same length , @xmath42 , where @xmath43 is the point of @xmath36 corresponding to @xmath2 and @xmath44 stands for the circumcenter ( see figure [ f1 ] ) .
there is only a finite number of ways to unfold sequences of faces , inducing a finite number of pairs of isometries @xmath45 , leading to a finite number of maps @xmath46 @xmath47 . for each zone
@xmath16 the function @xmath18 is either one @xmath48 or a constant map whose value is a vertex .
let @xmath49 be the square of the distance between @xmath2 and @xmath50 .
the algebraic arcs that compose @xmath51 are parts of the locus of those points such that @xmath52 for some indices @xmath53 and @xmath54 , such that @xmath55 for some index @xmath53 and some vertex @xmath14 , or such that @xmath56 for some vertices @xmath14 and @xmath57 .
a straightforward computation shows that they have degree at most @xmath12 .
( note that , in the only case that have been explicitly solved the regular tetrahedron the higher degree terms vanish and the actual degree drops to @xmath58 @xcite .
so it is not entirely clear that @xmath12 can be achieved . ) for further reasoning , it is necessary to notice that @xmath59 and @xmath60 can not be both translations ( lemma 1 in @xcite ) .
moreover , if either @xmath59 or @xmath60 is a translation then , by interchanging the roles of @xmath61 and @xmath62 , we obtain that @xmath60 and @xmath59 are both rotations , of the same angle .
hence , we can assume without loss of generality that @xmath45 are both rotations .
now , a straightforward computation proves the following lemma .
[ l1 ] let @xmath5 be a convex polyhedron , let @xmath16 be a zone of @xmath5 , and assume that @xmath18 is not constant .
endow the plane of @xmath16 and the plane of @xmath63 with orthonormal frames and let @xmath64 be the coordinates of @xmath65 .
the coordinates of @xmath3 are given by the formula:@xmath66 where @xmath67 , @xmath68 , @xmath69 , @xmath70 $ ] , @xmath71 , @xmath72 , @xmath73 .
moreover , the zero set of the denominator is neither empty nor reduced to a single point . indeed , @xmath74 if and only if the function @xmath45 are rotations of the same angle . in this case , the zero set of the denominator of ( [ 3 ] ) is the line through the centers of @xmath59 and @xmath60 . if the angles of the rotations are distinct ( , @xmath75 ) , the zero set of the denominator is the circle through the centers of @xmath59 , @xmath60 , and @xmath76 .
[ l2]with the notation of lemma [ l1 ] , in the case @xmath74 , there exists orthonormal frames of the planes of @xmath16 and @xmath77 such that @xmath78 let @xmath79 be the angle common to both rotations .
choose the orthonormal direct frame in the plane of @xmath63 in such a way that the coordinates of the center of @xmath41 are @xmath80 ( @xmath40 ) .
in the plane of @xmath16 , choose the frame in which the coordinates of @xmath81 equal the coordinates of @xmath82 .
let @xmath64 be the coordinates of @xmath2 ; the coordinates of @xmath83 are @xmath84 the equation in coordinates @xmath85 of the mediator of the segment @xmath86 is@xmath87 in order to shorten the formulas , we put @xmath88 and @xmath89 . the half difference and half sum of the two equations given by ( [ 4 ] ) ( @xmath40 ) are respectively@xmath90{rrr}u\xi & + v\psi & = -ux+vy\\ \left ( ux - vy\right ) \xi & + \left ( uy+vx\right ) \psi & = -u\text{.\hspace { 0.9cm}}\end{array}\ ] ] it follows that @xmath91@xmath92
now , we are in a position to prove the claimed result .
the idea of proof is very simple : we just have to prove that the inverse of a function of the form ( [ 3 ] ) can not be of this form .
any convex polyhedron @xmath5 contains an open and dense set @xmath93 such that @xmath94 is single - valued , @xmath95 is single - valued , and @xmath96 for all @xmath97 .
consequently , no polyhedron satisfies , even locally , steinhaus conditions .
it is sufficient to prove that any open set @xmath98 contains an open subset @xmath99 such that @xmath100 and @xmath101 are single - valued , and such that @xmath96 for any @xmath102 .
then , the union of all those @xmath99 is a suitable @xmath93 .
let @xmath103 be a fixed open set ; by virtue of theorem [ t0 ] , it contains a subset @xmath104 which is included in a zone .
since @xmath105 is continuous , there exists a smaller open set @xmath106 such that @xmath107 is wholly contained in a zone too . since @xmath108 is single - valued and thereby continuous , the set @xmath99 of those points in @xmath109 such that @xmath110 is open .
if it is non - empty , the proof is over ; suppose on the contrary that @xmath111 for all @xmath112 .
put @xmath113 .
let @xmath114 be the plane of the face containing @xmath115 and @xmath116 be the plane of the face containing @xmath117 .
by lemma [ l1 ] , @xmath118 and @xmath119 are rational and admit natural continuations @xmath120 and @xmath121 respectively , where @xmath122 and @xmath123 are either a circle or a line . by hypothesis , @xmath124 ; since @xmath125 and @xmath126 are rational , @xmath127 wherever @xmath128 is well defined .
assume that @xmath129 intersects @xmath123 .
let @xmath130 be a point of the boundary of @xmath131 , and @xmath132 be a point tending to @xmath130 when @xmath133 tends to infinity
. then @xmath134 tends to infinity , in contradiction with @xmath135 .
hence @xmath129 does not intersect @xmath123 , and similarly @xmath136 does not intersect @xmath122 .
assume now that either @xmath122 or @xmath123 is a circle .
since @xmath125 and @xmath126 play symmetrical roles , we can assume that @xmath122 is a circle .
let @xmath93 be the interior of @xmath122 and @xmath137 its exterior ; let @xmath138 , @xmath139 be the two connected components of @xmath140 . since @xmath141 tends to infinity when @xmath142 tends to some points of @xmath123 , @xmath143 ( resp .
@xmath144 ) can not be included in @xmath93 .
it follows that @xmath145 .
this contradict the fact that a point @xmath97 equals @xmath146 .
now assume that @xmath122 and @xmath123 are both straight lines .
by lemma [ l2 ] , one can assume that the expression of @xmath125 in euclidean coordinates is given by ( [ 3 ] ) and ( [ 2 ] ) .
let @xmath147 be the point of coordinates @xmath148 , @xmath149 be the point of coordinates @xmath150 and @xmath151 the point of coordinates @xmath152 .
a direct computation shows that@xmath153 hence @xmath154 and we get a contradiction .
in order to close the paper , we will mention a few open questions related to the above result .
the first two were already suggested in the introduction .
observe that the definition of steinhaus surfaces combines two distinct conditions : the fact that @xmath1 is single - valued does not imply that it is an involution .
the simplest known example is an ellipsoid of revolution whose axes measure respectively @xmath35 , @xmath35 , and @xmath155 . on such an ellipsoid , @xmath1 is a homeomorphism , but not an involution @xcite .
this naturally leads to the following question .
a certain trend nowadays is to consider the so - called _ alexandrov surfaces with curvature bounded below _ , instead of the classical convex surfaces .
roughly speaking , the difference between these notions is twofold . on the one hand ,
the curvature bound is no longer necessarily zero , and on the other hand , the topology is no longer spherical .
see @xcite for details . in this context , it is natural to consider _ abstract polyhedra _ obtained by gluing several polygons along their boundaries , in such a way that ( 1 ) the gluing map is length preserving , ( 2 ) the resulting space is a ( not necessarily spherical ) topological surface , and ( 3 ) the singular curvature at each vertex is non - negative . it is easy to see that the theorem of this paper can not be generalized to such abstract polyhedra , for rectangle flat tori are steinhaus . as for projective planes ,
the main obstruction to be steinhaus is purely topological .
assume that @xmath1 is single - valued , and consequently continuous .
since , as a consequence of lefschetz fixed point theorem , even dimensional projective spaces have the fixed point property , @xmath1 should have a fixed point , which is absurd .
still in the context of alexandrov surfaces , it is natural to consider abstract polyhedra whose faces are geodesic polygons of the unit sphere ( _ spherical polyhedra _ ) , or of the hyperbolic plane ( _ hyperbolic polyhedra _ ) . | on a convex surface @xmath0 , the antipodal map @xmath1 associates to a point @xmath2 the set of farthest points from @xmath2 , with respect to the intrinsic metric .
@xmath0 is called a steinhaus surface if @xmath1 is a single - valued involution .
we prove that any convex polyhedron has an open and dense set of points @xmath2 admitting a unique antipode @xmath3 , which in turn admits a unique antipode @xmath4 , distinct from @xmath2 .
in particular , no convex polyhedron is steinhaus .
keywords : convex polyhedra , antipodal map .
msc 2010 : 52b10 , 51a15 , 53c45 |
crystallization , during which a material transforms from a dense , amorphous liquid to a crystalline solid , occurs via the nucleation and subsequent growth of small crystalline domains @xcite .
crystallization in metals has been intensely studied over the past several decades with the goal of developing the ability to tune the microstructure to optimize the mechanical properties of metal alloys @xcite . however , in - situ observation of crystallization in metallic melts is limited due to the rapid crystallization kinetics of metals @xcite .
in contrast , bulk metallic glasses ( bmgs ) , which are amorphous metal alloys , can be supercooled to temperatures below the solidus temperature @xmath10 and persist in a dense , amorphous liquid state over more than @xmath11 orders of magnitude in time scales or viscosity @xcite .
deep supercooling of bmgs provides the ability to study crystallization on time scales that are accessible to experiments @xcite .
these prior experimental studies have uncovered fundamental questions concerning crystallization kinetics in bmgs .
for example ,
when a bmg in the glass state is heated to a temperature @xmath12 in the supercooled liquid region , crystallization is much faster than crystallization that occurs when the metastable melt is cooled to the same temperature @xmath13 @xcite .
asymmetries in the crystallization time scales upon heating versus cooling of up to two orders of magnitude have been reported in experiments @xcite .
the asymmetry impacts industrial applications of bmgs because rapid crystallization upon heating limits the thermoplastic forming processing time window for bmgs @xcite .
recent studies have suggested that the asymmetry in the crystallization time scales originates from the temperature dependence of the the nucleation and growth rates @xcite , _
i.e. _ that the nucleation rate is maximal at a temperature below that at which the growth rate is maximal .
according to this argument , crystallization upon heating is faster because of the growth of the nascent crystal nuclei that formed during the thermal quench to the glass .
in contrast , crystallization is slower upon cooling since crystal nuclei are not able to form at high temperatures in the melt .
however , there has been no direct visualization of the crystallization process in bmgs , and it is not yet understood why the asymmetry varies from one bmg to another @xcite and how sensitively the asymmetry depends on the cooling rate @xmath2 used to prepare the glass .
an improved , predictive understanding of the crystallization process in bmgs will aid the design of new bmg - forming alloys with small crystallization asymmetry ratios and large thermoplastic processing time windows .
@xmath14{h1.eps } & \includegraphics[width = 1.0 in]{h2.eps } & \includegraphics[width = 1.0 in]{h3.eps } & \includegraphics[width = 1.0 in]{h4.eps } & \includegraphics[width = 1.0 in]{h5.eps } & \includegraphics[width = 1.2 in]{nclusterh.eps } \\ \includegraphics[width = 1.0 in]{c1.eps } & \includegraphics[width = 1.0 in]{c2.eps } & \includegraphics[width = 1.0 in]{c3.eps } & \includegraphics[width = 1.0 in]{c4.eps } & \includegraphics[width = 1.0 in]{c5.eps } & \includegraphics[width = 1.2 in]{nclusterc.eps } \\ \end{array}$ ] we employ molecular dynamics ( md ) simulations of bidisperse spheres interacting via lennard - jones potentials @xcite to visualize directly the crystallization process upon heating and cooling in model metallic glass - forming systems .
we perform thermal quenches of the system from a high temperature @xmath15 in the equilibrated liquid regime to a glass at @xmath16 and vary the cooling rate @xmath1 by several orders of magnitude .
for cooling rates below the critical cooling rate @xmath17 , the system begins to crystallize , whereas for @xmath18 , the system remains amorphous .
we also performed md simulations in which we heat the zero - temperature glassy states ( prepared at cooling rate @xmath19 ) through the supercooled liquid regime to @xmath20 over a range of heating rates @xmath4 . for heating rates @xmath21 ,
the system begins to crystallize , whereas for @xmath22 , it remains amorphous .
we also find that the critical heating rate has an intrinsic contribution @xmath23 and an @xmath2-dependent contribution @xmath24 that increases with decreasing @xmath2 .
we measured the asymmetry ratio @xmath8 as a function of the glass - forming ability ( gfa ) and @xmath2 for several binary lennard - jones mixtures and find that @xmath25 and the ratio grows with increasing gfa and decreasing @xmath2 .
we show that these results are consistent with predictions from classical nucleation theory ( cnt ) that the maximal growth rate occurs at a higher temperature than the maximal nucleation rate and that the separation between the nucleation and growth peaks increases with the gfa .
further , cnt is able to qualitatively recapitulate the dependence of the asymmetry ratio on the gfa as measured through @xmath6 for both our md simulations and recent experiments on bmgs as well as on @xmath2 for the md simulations @xcite .
the remainder of the manuscript is organized into three sections : sec .
[ sec : method ] : methods , sec .
[ sec : results ] : results , and sec .
[ sec : discussion ] : conclusion . in sec .
[ sec : method ] , we describe the md simulations of binary lennard - jones mixtures , the computational methods to detect and structurally characterize crystal nuclei , and measurements of the critical cooling and heating rates , @xmath6 and @xmath5 . in sec .
[ sec : results ] , we show results from md simulations for the time - temperature transformation diagram @xcite and the asymmetry ratio @xmath8 as a function of the glass - forming ability as measured by the critical cooling rate @xmath6 and the cooling rate used to prepare the zero - temperature glasses @xmath2 .
we also compare our results for the asymmetry ratio to experimental measurements of the ratio for two bmgs and to predictions of the ratio from classical nucleation theory . in sec .
[ sec : discussion ] , we briefly summarize our results and put forward our conclusions .
we performed md simulations of binary lennard - jones ( lj ) mixtures of @xmath26 spheres with mass @xmath27 at constant volume @xmath28 in a cubic simulation box with side length @xmath29 and periodic boundary conditions .
we studied mixtures with @xmath30 and diameter ratio @xmath31 .
we employed the lj pairwise interaction potential between spheres @xmath32 and @xmath33 : @xmath34,\ ] ] where @xmath35 is their center - to - center separation , @xmath36 is the depth of the minimum in the potential energy @xmath37 , @xmath38 , and @xmath37 has been truncated and shifted so that the potential energy and force vanish for separations @xmath39 @xcite .
we varied the system volume @xmath40 to fix the packing fraction @xmath41 @xcite at each diameter ratio @xmath42 .
for most simulations , we considered @xmath43 spheres , but we also studied @xmath44 and @xmath45 to assess finite - size effects .
below , energy , length , time , and temperature scales are expressed in units of @xmath36 , @xmath46 , @xmath47 , and @xmath48 , respectively , where the boltzmann constant @xmath49 has been set to be unity . for each particle diameter ratio , which yield different glass - forming abilities ,
we performed md simulations to cool metastable liquids to zero temperature and heat zero - temperature glasses into the metastable liquid regime to measure @xmath6 and @xmath5 at which the systems begin to crystallize .
to measure @xmath6 , we first equilibrate the system at high temperature @xmath50 using a gaussian constraint thermostat @xcite .
we then cool the system by decreasing the temperature linearly at rate @xmath1 from @xmath15 to @xmath16 : @xmath51 to measure the critical heating rate @xmath52 at finite rate @xmath2 , we first prepare the systems in a glass state by cooling them from the high temperature liquid state to zero - temperature at rate @xmath53 . to measure the intrinsic critical heating rate @xmath23 , we quench the systems infinitely fast to zero temperature using conjugate gradient energy minimization . for both cases , we heat the zero - temperature glasses using a linear ramp @xmath54 until @xmath55 . for both heating and cooling protocols , we carried out @xmath56 independent trajectories and averaged the results .
to detect the onset of crystallization in our simulations @xcite , we differentiate ` crystal - like ' versus ` liquid - like ' particles based on the value of the area - weighted bond orientational order parameter for each particle @xcite . we define the complex - valued bond orientational order parameter for particle @xmath32 : @xmath57 where @xmath58 is the spherical harmonic of degree @xmath59 and order @xmath27 , @xmath60 and @xmath61 are the polar and azimuthal angles for the vector @xmath62 , @xmath63 gives the index of the voronoi neighbors of particle @xmath32 , and @xmath64 is the area of the face of the voronoi polyhedron common to particles @xmath32 and @xmath33 .
the correlation coefficient @xcite between the bond orientational order parameters @xmath65 and @xmath66 , where particle @xmath33 is a voronoi neighbor of @xmath32 , @xmath67 is sensitive to face - centered - cubic ( fcc ) order .
when @xmath68 , @xmath32 and @xmath33 are considered ` connected ' . if particle @xmath32 has more than @xmath69 connected voronoi neighbors , it is defined as ` crystal - like ' .
the ratio @xmath70 gives the fraction of crystal - like particles in a given configuration .
in addition , we also define a crystal cluster as the set of crystal - like particles that possess mutual voronoi neighbors .
distinct crystal clusters that nucleate and grow upon heating and cooling are shown in fig . [
fig : snapshot ] .
@xmath71{allin.eps } \end{array}$ ] for each diameter ratio and rate , we measure the probability for crystallization @xmath72 , where @xmath73 is the number of trajectories that crystallized with @xmath74 during the heating or cooling protocol and @xmath75 is the total number of trajectories ( _ cf . _
insets to fig .
[ fig : coolheat ] ) .
we find that the data for @xmath76 collapses onto a sigmoidal scaling function as shown in fig .
[ fig : coolheat ] : @xmath77,\ ] ] where @xmath78 is the probability for crystallization in the limit of infinitely fast rates @xmath79 , @xmath80 is the probability for crystallization in the @xmath81 limit , @xmath82 is the rate at which @xmath83 , and @xmath84 is the stretching factor .
we find that @xmath85 and @xmath86 for @xmath87 , and these factors increase by only a few percent over the range in @xmath42 that we consider .
we define the critical heating and cooling rates @xmath88 and @xmath89 by the rates at which @xmath90 , _ i.e. _ @xmath91}.\ ] ] as shown in the insets to fig .
[ fig : coolheat ] , for @xmath92 most of the configurations crystallize during heating or cooling .
in contrast , for @xmath93 , none of the configurations crystallize .
an advantage of md simulations is that they can provide atomic - level structural details of the crystallization dynamics that are often difficult to obtain in experiments . in fig .
[ fig : snapshot ] , we visualize the nucleation and growth of clusters of crystal - like particles during the heating and cooling simulations . in both cases , the number of clusters reaches a maximum near @xmath94 . in fig .
[ fig : nclusters ] , we show the maximum number of clusters @xmath95 ( normalized by @xmath96 ) that form during the heating and cooling protocols .
we find that more crystal clusters form during the heating protocol compared to the cooling protocol for all particle diameter ratios studied , which is supported by the measured time - temperature - transformation ( ttt ) diagram .
in addition , we will show below that the asymmetry ratio @xmath97 , and that the ratio grows with increasing gfa ( increasing diameter ratio ) and decreasing @xmath2 .
we find that cnt can qualitatively describe the dependence of the asymmetry ratio on the gfa , as measured by the critical cooling rate @xmath6 , for both our md simulations and recent experiments on bmgs , as well as on the preparation cooling rate @xmath2 for the md simulations .
the critical heating and cooling rates can be obtained by fitting the probability for crystallization @xmath76 as a function of @xmath4 or @xmath1 to the sigmoidal form in eq .
[ pcrystal ] .
we first investigate the minimum value for the asymmetry ratio @xmath98 , which is obtained by taking the @xmath99 limit .
( the asymmetry ratio @xmath100 for finite preparation rates @xmath2 will be considered in sec .
[ rp ] . ) in fig .
[ fig : rceffect ] , we plot @xmath98 versus @xmath89 ( for diameter ratios @xmath101 , @xmath102 , @xmath103 , @xmath104 , and @xmath105 ) .
we find that @xmath106 for all systems studied , which is consistent with classical nucleation theory ( cnt ) .
as shown in fig .
[ fig : snapshot ] , more crystal nuclei form during the heating protocol than during the cooling protocol .
in addition , cnt predicts that the growth rates for crystal nuclei are larger during heating compared to cooling . in sec .
[ cnt_prediction ] , we will show that both factors contribute to an increased probability for crystallization during heating . in fig .
[ fig : rceffect ] , we also show that the asymmetry ratio @xmath107 increases as the critical cooling rate @xmath6 decreases , or equivalently as the glass - forming ability increases . in the md simulations
, we were able to show a correlation between the asymmetry ratio and the critical cooling rate over roughly an order of magnitude in @xmath6 . in sec .
[ cnt_prediction ] , we introduce a model that describes qualitatively this dependence of the asymmetry ratio on @xmath6 . in classical nucleation theory ( cnt ) , the formation of crystals is a two - step process : 1 ) fluctuations in the size of crystal nuclei that allow them to reach the critical radius @xmath108 and 2 ) growth of post - critical nuclei with @xmath109 . to form a critical nucleus ,
the system must overcome a nucleation free energy barrier : @xmath110 where @xmath111 is the bulk gibbs free energy difference per volume ( in units of @xmath112 ) and @xmath113 is the surface tension between the solid and liquid phases ( in units of @xmath114 ) .
we assume that @xmath115 @xcite , where @xmath116 is melting temperature , @xmath117 is the degree of undercooling , and @xmath118 is a dimensionless parameter that characterizes the thermodynamic drive to crystallize and will be used to tune the gfa of the system ( where @xmath119 is the latent heat of fusion ) . within cnt
, the rate of formation of critical nuclei ( nucleation rate ) is given by : @xmath120 where @xmath121 is an @xmath122 constant with units @xmath123 , @xmath124 is the atomic diffusivity with units @xmath125 , and @xmath126 is an effective activation energy for the diffusivity ( with units of @xmath36 ) . after the nucleation free energy barrier @xmath127 has been overcome and crystal nuclei reach @xmath128 , the growth rate of crystal nuclei is given by @xmath129,\ ] ] where @xmath130 the characteristic interatomic spacing . in fig . [
fig : iuschematic ] , we plot the nucleation @xmath131 and growth rates @xmath132 with @xmath133 and @xmath134 from md simulations of binary lj systems @xcite , @xmath135 , which is typical for bmgs @xcite , while varying the gfa parameter from @xmath136 to @xmath137 ( corresponding to diameter ratios from @xmath101 to @xmath105 . )
both @xmath138 and @xmath139 are peaked with maxima @xmath140 and @xmath141 at temperatures @xmath142 and @xmath143 . in fig .
[ fig : iuschematic ] , we show that as the gfa increases , @xmath140 and @xmath141 , as well as @xmath142 and @xmath143 decrease
. however , @xmath142 decreases faster than @xmath143 , so that the separation between the peaks , @xmath144 , increases with gfa . to determine the critical heating and cooling rates , @xmath5 and @xmath6
, we must calculate the fraction of the samples @xmath73 that crystallize and the probability for crystallizing @xmath145 , where @xmath75 is the total number of samples , upon heating and cooling .
within classical nucleation theory , the probability to crystallize upon cooling from @xmath15 to @xmath13 is given by @xcite : @xmath146 ^ 3 dt ' .
\label{eq : x}\ ] ] we assume that @xmath15 is above the liquidus temperature @xmath0 , and @xmath13 is below the glass transition temperature @xmath147 , where the time required to form crystal nuclei diverges .
we can rearrange eq .
[ eq : x ] to solve for the critical cooling rate at which @xmath148 : @xmath149 ^ 3 dt ' \\ & = a ' \int^{t_f}_{t_i } dt ' \exp\left(-\frac{q_{eff}}{t'}\right ) \exp\left(-\frac{\delta g^{\ast}}{t'}\right ) \\ &
\left[\int^{t_f}_{t ' } \exp\left(-\frac{q_{eff}}{t''}\right ) \left[1 - \exp\left(-\frac{\delta g v}{t''}\right)\right ] dt''\right]^3 , \label{eq : rc } \end{split}\ ] ] where @xmath150 and we assumed that @xmath121 , @xmath124 , and @xmath130 are independent of temperature .
a similar expression for the intrinsic critical heating rate @xmath23 can be obtained by reversing the bounds of integration in eq .
[ eq : rc ] . in fig .
[ fig : rceffect ] , we plot the intrinsic asymmetry ratio @xmath151 predicted from eq .
[ eq : rc ] versus the critical cooling rate @xmath6 after choosing the best value @xmath152 that interpolates between the md simulation data at high @xmath6 and experimental data from bmgs at low @xmath6 .
we find that cnt qualitatively captures the increase in the asymmetry ratio with increasing gfa over a wide range of critical cooling rates from @xmath153 ( experiments on bmgs ) to @xmath154 ( md simulations of binary lj systems ) .
a comparison of figs .
[ fig : rceffect ] and [ fig : iuschematic ] reveals that the increase in the intrinsic asymmetry ratio is caused by the separation of the peaks in the growth and nucleation rates @xmath139 and @xmath138 that occurs as the gfa increases .
thus , we predict an enhanced value for @xmath144 in experiments on bmgs since the critical cooling rate in experiments is orders of magnitude smaller than in the md simulations .
the fact that @xmath155 is also reflected in the asymmetry of the `` nose '' of the time - temperature - transformation ( ttt ) diagram . in fig
[ fig : ttt ] , we show the probability @xmath156 that the system has crystallized at a given temperature @xmath157 after a waiting time @xmath158 for monodisperse lj systems .
we find that @xmath159-@xmath160 is the temperature at which the waiting time for crystallization is minimized and that the time to crystallize is in general longer for @xmath161 compared to @xmath162 .
because crystallization on average occurs at a higher temperature during heating and a lower temperature during cooling , the asymmetry in the ttt diagram indicates that slower rates are required to crystallize during cooling than during heating ,
_ i.e. _ @xmath163 . in sec .
[ cnt_prediction ] , we assumed that the initial samples ( _ i.e. _ the zero - temperature glasses ) for the heating protocol were prepared in the @xmath164 limit and , thus were purely amorphous .
how does the asymmetry ratio @xmath165 depend on @xmath2 when the preparation cooling rate @xmath2 is finite and partial crystalline order can occur in the samples ? in this section , we show results for the asymmetry ratio @xmath165 for monodisperse systems using a protocol where the samples are quenched from equilibrated liquid states to zero temperature at a finite rate @xmath2 and then heated to temperature @xmath13 at rate @xmath4 .
( see sec .
[ protocol ] . )
note that when @xmath166 , some of the samples crystallize during the cooling preparation , yet these samples are still included in the calculation of the probability @xmath167 to crystallize . in fig .
[ fig : rpeffect ] , we show the results for the asymmetry ratio @xmath165 from md simulations . we find that @xmath165 grows rapidly as @xmath2 approaches @xmath6 from above and reaches a plateau value of @xmath168 in the limit @xmath169 . the critical heating rate @xmath52 at finite @xmath2 can also be calculated from cnt using an expression similar to eq .
[ eq : rc ] with an additional term that accounts for cooling the equilibrated liquid samples to zero temperature at a finite rate . in fig .
[ fig : rpeffect ] , we show that the asymmetry ratio @xmath165 predicted using cnt agrees qualitatively with that from the md simulations . the number of crystal nuclei that form during the quench increases with decreasing @xmath2 , which causes @xmath165 to diverge as @xmath9 .
the predicted _ intrinsic _ contribution to the asymmetry ratio for @xmath170 is small , and @xmath165 is dominated by the preparation protocol . in contrast , the asymmetry ratio @xmath171 is dominated by the intrinsic contribution in the @xmath172 limit . as shown in fig . [ fig : rceffect ] , the size of the intrinsic contribution to the asymmetry ratio can be tuned by varying the gfa , which controls the separation between the peaks in the nucleation @xmath138 and growth @xmath139 rates .
we performed md simulations of binary lennard - jones systems to model the crystallization process during heating and cooling protocols in metallic glasses .
we focused on measurements of the ratio of the critical heating @xmath5 and cooling @xmath6 rates , below which crystallization occurs during the heating and cooling trajectories .
we find : 1 ) @xmath7 for all systems studied , 2 ) the asymmetry ratio @xmath8 grows with increasing glass - forming ability ( gfa ) , and 3 ) the critical heating rate @xmath52 has an intrinsic contribution @xmath23 and protocol - dependent contribution @xmath24 that increases with decreasing cooling rates @xmath2 used to prepare the initial samples at zero temperature .
we show that these results are consistent with the prediction from classical nucleation theory that the maximal growth rate occurs at a higher temperature than the maximal nucleation rate and that the separation between the peaks in nucleation @xmath138 and growth @xmath139 rates increases with the gfa .
predictions from cnt are able to qualitatively capture the dependence of the asymmetry ratio on the gfa as measured through @xmath6 for both our md simulations and recent experiments on bmgs as well as on @xmath2 for the md simulations .
thus , our simulations have addressed how the thermal processing history affects crystallization , which strongly influences the thermoplastic formability of metallic glasses .
the authors acknowledge primary financial support from the nsf mrsec dmr-1119826 .
we also acknowledge support from the kavli institute for theoretical physics ( through nsf grant no .
phy-1125915 ) , where some of this work was performed .
this work also benefited from the facilities and staff of the yale university faculty of arts and sciences high performance computing center and the nsf ( grant no .
cns-0821132 ) that in part funded acquisition of the computational facilities . | we perform molecular dynamics ( md ) simulations of the crystallization process in binary lennard - jones systems during heating and cooling to investigate atomic - scale crystallization kinetics in glass - forming materials . for the cooling protocol , we prepared equilibrated liquids above the liquidus temperature @xmath0 and cooled each sample to zero temperature at rate @xmath1 . for the heating protocol , we first cooled equilibrated liquids to zero temperature at rate @xmath2 and then heated the samples to temperature @xmath3 at rate @xmath4 .
we measured the critical heating and cooling rates @xmath5 and @xmath6 , below which the systems begin to form a substantial fraction of crystalline clusters during the heating and cooling protocols .
we show that @xmath7 , and that the asymmetry ratio @xmath8 includes an intrinsic contribution that increases with the glass - forming ability ( gfa ) of the system and a preparation - rate dependent contribution that increases strongly as @xmath9 from above .
we also show that the predictions from classical nucleation theory ( cnt ) can qualitatively describe the dependence of the asymmetry ratio on the gfa and preparation rate @xmath2 from the md simulations and results for the asymmetry ratio measured in zr- and au - based bulk metallic glasses ( bmg ) .
this work emphasizes the need for and benefits of an improved understanding of crystallization processes in bmgs and other glass - forming systems . |
nearly all calculations for the determination of mixing angles between different spin - components of baryon wavefunctions are based on spectroscopic models . beside their advantages
these models also contain some problems which are not dissolved yet .
+ the goal of the presented work is an independent way of the determination of mixing angles and range parameters in the harmonic oscillator description of nucleonic resonances within the nonrelativistic quark model .
although the approach in a first step is applied to the spin1/23/2mixing angle @xmath2 in the negative parity resonance @xmath3 ( the underlying nature of this resonance is at the moment object to various speculations ) , the approach also can be applied to many other resonances .
+ the idea of the presented method is as follows : in a first step one has to calculate the total cross section of a process which is dominated by the resonance under consideration in such a way that the only remaining free parameters of the calculation are the mixing angles and the range parameters of the harmonic oscillator wavefunctions of the resonance and the involved ground state nucleons . in a second step
the parameters are obtained by adjusting the calculated to the experimentally measured cross section .
+ one process dominated mainly by the @xmath3 is the reaction @xmath1 at threshold . in order to determine the
desired mixing parameters one first has to get a _
quantitative _ understanding not only of the _ short range dynamics _ of the process , but also of the _ initial and final state interactions _ ( isi and fsi ) between the in- and outgoing particles .
main investigations along this line have been performed in @xmath4@xcite and references therein .
the reaction @xmath1 at threshold is described within a relativistic meson - exchange model .
the virtual @xmath3 is excited / deexited by the exchange of @xmath5 , @xmath6 , @xmath7 , @xmath8 , @xmath9 and @xmath10mesons , while it is deexcited / excited by the produced @xmath8 .
using relativistic normalisations for spinors , creation and annihilation operators it is straight forward to write down the expression for the total cross section of @xmath1 ( @xmath11 , @xmath12 , @xmath13 , @xmath14 ) : @xmath15 the combinatorial factor of @xmath16 is due to the two outgoing identical protons . for the relativistic description of @xmath1 the following 5 independent invariants are chosen : @xmath17 in order to get a _
quantitative _ understanding of the isi and fsi it is crucial to recall the steps which lead to that what the people nowadays callwatson - migdal approximation " and what is wrongly implemented in nearly all threshold meson production calculations by now : the system of colliding and produced particles is described by an overall hamilton operator @xmath18 , while @xmath19 is the kinetic energy operator , @xmath20 is the particle number nonconserving short range interaction potential and @xmath21 is the particle number conserving interaction potential of long range . for convenience
additionally one can define the long range hamilton operator @xmath22 .
the corresponding eigenstates to the operators @xmath23 , @xmath24 and @xmath19 fulfil the following schrdinger equations : @xmath25 the corresponding lippmann
schwinger equations are : @xmath26 defining the free propagator @xmath27 and inserting complete sets of free asymptotic states the t - matrix @xmath28 in eq .
( 1 ) can be separated in a short and long ranged part via ( @xmath29 ) : @xmath30 as the in- and outgoing particles in @xmath1 are nearly on - shell one can approximate the free propagator @xmath31 in the following way : @xmath32 at this point it is useful to introduce the following relative lab wavenumbers for each pair of particles in the initial and final state ( @xmath33 ) : @xmath34 and the corresponding wavevectors @xmath35 , @xmath36 , @xmath37 , @xmath38 .
consider now the case , when the only two particles taking part in the isi and fsi are the two in and outgoing protons . in this case
the subsystem of the two protons interacting via the long range potental @xmath21 with phaseshifts @xmath39 ( @xmath40 = orbital angular momentum ) can be treated within the framework of standard nonrelativistic scattering theory . by application of the on - shell approximation of eq .
( 6 ) the t - matrix in eq .
( 5 ) can be transformed to ( @xmath41 , @xmath42 ) : @xmath43 here the jost functions @xmath44 were introduced by @xmath45 .
it is worth to mention that the isi and fsi factors of eq .
( 8) have the correct limit for vanishing isi or fsi .
if there is no isi or fsi , the phaseshifts will vanish and the corresponding factors will go to 1 . the effect of the watson - migdal approach is a factorization of @xmath28 into a short ranged t - matrix @xmath46 and isi- and fsi - factors @xmath47 and @xmath48 .
the the factor @xmath47 is only a function of @xmath49 , while the short ranged t - matrix @xmath46 close to threshold shows up to be a slowly varying function of the phasespace integration variables and therefore can be set to its threshold value @xmath50 . using this observation
both factors can be drawn in front of the phasespace integral in eq .
( 1 ) with the result : @xmath51 in the last line of eq . ( 9 ) the final state interaction modified phasespace integral @xmath52 for the three body final state @xmath53 has been defined . to expand
the energy dependence of the total cross section at threshold one usually expresses the cross section in terms of the dimesionless variable @xmath54 instead of the square of the cm - energy @xmath49 .
the quantity @xmath8 being the maximum momentum of the produced @xmath8-meson in the cm - frame vanishes at threshold .
it is easy to derive @xmath55 . introducing @xmath56 and @xmath57 the phaseshifts of the long range potential between the incoming two protons
can be expanded at threshold : @xmath58 in terms of these expansion coefficents the factor @xmath47 can be expanded at threshold : @xmath59 at the moment there is no clear knowledge of the values of @xmath60 and @xmath61 at the production threshold of @xmath1 , but as the incoming relative proton momentum is very large the interaction time is very short , so that one might assume @xmath60 and @xmath61 to be zero which yields @xmath62 , while the vpi - values @xmath63 , @xmath64 lead to @xmath65 .
+ for the complete determination of @xmath48 one obviously has to solve the faddeev equations for the outgoing @xmath53system . as this is at the moment out of the scope of this work only the leading terms of the fadeev expansion
are taken into account : @xmath66 the various terms in eq .
( 12 ) which denote the fsi between each pair of particles in the final state can be expanded by effective range expansions . for the outgoing pp
system the s - wave nuclear effective range expansion is given by : @xmath67 taking into account only pp - fsi using the shape independent effective range expansion eq .
( 13 ) the fsi - modified phasespace integral can be expanded at threshold ( @xmath68 ) : @xmath69 + o ( \eta^8 ) \right\}\end{aligned}\ ] ] it is interesting to observe that even the @xmath70-term is independent of the effective range @xmath71 . for the relativistic meson - exchange model
the following compact expression for the @xmath72 threshold transition amplitude @xmath73 of @xmath1 is obtained @xmath4@xcite ( @xmath74 ) : @xmath75 \ , - \nonumber \\ & - & x_\rho \ ; m_{\scriptscriptstyle n } \ ; [ \ , 4 \ , ( m_\eta - 2\ , m_{\scriptscriptstyle n } ) + k_\rho \ , ( 5 m_\eta - 4\ , m_{\scriptscriptstyle n } ) \ , ] \ ; + \nonumber \\ & + & [ \ , y_{\,\rho } \ , ( m_{\scriptscriptstyle n } + m_\eta ) + \tilde{m}_\rho \ , ] \ ; [ \ , k_{\rho } \ ; ( m_\eta - 4\ , m_{\scriptscriptstyle n } ) - 8\ , m_{\scriptscriptstyle n } \ , ] - \nonumber \\ & - & x_\omega \ ; m_{\scriptscriptstyle n } \ ; \left [ \ , 4 \ , ( m_\eta - 2\ , m_{\scriptscriptstyle n } ) + k_\omega \ , ( 5 m_\eta - 4\ , m_{\scriptscriptstyle n } ) \ , \right ] \ ; + \nonumber \\ & + & [ \ , y_{\,\omega } \ , ( m_{\scriptscriptstyle n } + m_\eta ) + \tilde{m}_\omega \ , ] \ ; [ \ , k_{\omega } \ ; ( m_\eta - 4\ , m_{\scriptscriptstyle n } ) - 8\ , m_{\scriptscriptstyle n } \ , ] \qquad ( k_\rho \approx 6.1 , k_\omega \approx 0)\end{aligned}\ ] ] here i used the following abbreviations ( @xmath76 ) ( @xmath77 ) ( @xmath78 ) ( @xmath79 , @xmath80 , @xmath81 ) : @xmath82 the meson
nucleon@xmath3 couplings are in general complex and show a strong nontrivial momentum sensitivity .
for that reason it is not enough to evaluate eq .
( 16 ) applying the commonly used on - shell coupling constants combined with standard monopole or dipole formfactors .
hence a model has been developed to estimate the real and imaginary parts of the couplings @xmath4@xcite .
the real parts of the couplings are derived within the framework of a nonrelativistic quark model , in which the ground state nucleon and the @xmath3 wavefunctions are described by harmonic oscillator solutions .
the radial wavefunctions @xmath83 , @xmath84 and @xmath85 of the ground state nucleon and the @xmath3 are determined by one unique range parameter @xmath86 ( here @xmath87 , @xmath88 are 3quark jacobian coordinates ! ) : @xmath89 the imaginary parts of the couplings are calculated close the threshold from relevant lowest order meson loop corrections to the bare couplings .
taking into account only @xmath7 , @xmath8 , @xmath9 and @xmath10exchange we observe by reproducing the experimental total cross section of @xmath90 that the following mixing parameters are favourable : @xmath91 , @xmath92 fm . because of the uncertainties in the bare meson - nucleon-@xmath3 couplings and the still improvable treatment of isi and fsi these numbers are by now preliminary .
the model gives for the first time a quantitative prediction of the relative phases between the @xmath7 , @xmath8 , @xmath9 and @xmath10exchange amplitudes .
an extension of the effective range formalism for isi and fsi to coulomb - interactions is on the way .
99 f. kleefeld , exklusive schwellenproduktion von @xmath93 , @xmath8 und @xmath94mesonen in proton proton sten " , doctoral thesis , university of erlangen ( 1999 ) .
f. kleefeld , m. dillig ,
protoninduced pion- , eta- and kaon - production at threshold " , in _ proc . of @xmath95 int .
conf . on the structure of baryons ( baryons 98 ) _
( bonn , sept . 22 - 26 , 1998 ; 1999 by world scientific publishing ) pp .
613616 , nucl - th/9811003 . f. kleefeld , m dillig , reinvestigation of @xmath96 and @xmath1 at threshold " , acta phys . pol . * b29 * , 3059 ( 1998 ) , nucl - th/9806057v2 . | a new method for the determination of the spin-1/2 - 3/2-mixing angle and the range parameter of the quarkmodel wavefunction of the resonance @xmath0 is presented .
the method ist based on a quantitative calculation of the total cross section of @xmath1 at threshold .
the quantitative on - shell treatment of isi and fsi is discussed . |
in his contribution to the 1962 international congress of mathematicians , igor shafarevich formulated an influential conjecture , considering smooth , projective families @xmath1 of curves of genus @xmath2 , over a fixed smooth quasi - projective base curve @xmath3 .
one part of the conjecture , known as the `` hyperbolicity conjecture '' , gives a sufficient criterion to guarantee that any such family is isotrivial .
the conjecture was shown in two seminal works by parshin and arakelov , including the following special case .
[ thm : shafarevich ] let @xmath1 be a smooth , complex , projective family of curves of genus @xmath2 , over a smooth quasi - projective base curve @xmath3 . if @xmath3 is isomorphic to one of the following varieties , * the projective line @xmath4 , * the affine line @xmath5 , * the affine line minus one point @xmath6 , or * an elliptic curve , then any two fibres of @xmath7 are necessarily isomorphic . throughout this paper , a _
family _ is a flat morphism of algebraic varieties with connected fibres .
we always work over the complex number field .
following standard convention , we refer to theorem [ thm : shafarevich ] as `` shafarevich hyperbolicity conjecture '' rather than `` arakelov - parshin theorem '' .
the reader interested in a complete picture is referred to @xcite , where all parts of the shafarevich conjecture are discussed in more detail .
formulated in modern terms , theorem [ thm : shafarevich ] asserts that any morphism from a smooth , quasi - projective curve @xmath3 to the moduli stack of algebraic curves is necessarily constant if @xmath3 is one of the special curves mentioned in the theorem . if @xmath3 is a quasi - projective variety of arbitrary dimension , then any morphism from @xmath3 to the moduli stack contracts all rational and elliptic curves , as well as all affine lines and @xmath6s that are contained in @xmath3 .
we refer to ( * ? ? ?
16.e.1 ) for a discussion of the relation between the shafarevich hyperbolicity conjecture and the notions of brody and kobayashi hyperbolicity .
this survey is concerned with generalisations of the shafarevich hyperbolicity conjecture to higher dimensions , concentrating on methods and results that relate moduli and minimal model theory .
we hope that the methods presented here will be applicable to a much wider ranges of problems , in moduli theory and elsewhere .
the list of problems that we would like to address include the following .
apart from the quasi - projective curves mentioned above , what other varieties admit only constant maps to the moduli stack of curves ?
what about moduli stacks of higher dimensional varieties ? given a variety @xmath3 , is there a good geometric description of the subvarieties that will always be contracted by any morphism to any reasonable moduli stack ?
much progress has been achieved in the last years and several of the questions can be answered today .
it turns out that there is a close connection between the minimal model program of a given quasi - projective variety @xmath3 , and its possible morphisms to moduli stacks .
some of the answers obtained are in fact consequences of this connection . in the limited number of pages available ,
we say almost nothing about the history of higher dimensional moduli , or about the large body of important works that approach the problem from other points of view .
hardly any serious attempt is made to give a comprehensive list of references , and the author apologises to all those whose works are not adequately represented here , be it because of the author s ignorance or simply because of lack of space .
the reader who is interested in a broader overview of higher dimensional moduli , its history , complete references , and perhaps also in rigidity questions for morphisms to moduli stacks is strongly encouraged to consult the excellent surveys found in this handbook and elsewhere , including @xcite .
a gentle and very readable introduction to moduli of higher dimensional varieties in also found in @xcite , while viehweg s book @xcite serves as a standard technical reference for the construction of moduli spaces .
most relevant notions and facts from minimal model theory can either be found in the introductory text @xcite , or in the extremely clear and well - written reference book @xcite .
recent progress in minimal model theory is surveyed in @xcite .
section [ sec : generalization ] introduces a number of conjectural generalisations of the shafarevich hyperbolicity conjecture and gives an overview of the results that have been obtained in this direction .
in particular , we mention results relating the moduli map and the minimal model program of the base of a family . sections [ sec : vz ] and [ sec : ext ] introduce the reader to methods that have been developed to attack the conjectures mentioned in section [ sec : generalization ] . while section [ sec : vz ] concentrates on positivity results on moduli spaces and on viehweg and zuo s construction of ( pluri-)differential forms on base manifolds of families , section [ sec : ext ] summarises results concerning differential forms on singular spaces .
both sections contain sketches of proofs which aim to give an idea of the methods that go into the proofs , and which might serve as a guideline to the original literature .
the introduction to section [ sec : ext ] motivates the results on differential forms by explaining a first strategy of proof for a special case of a ( conjectural ) generalisation of the shafarevich hyperbolicity conjecture . following this plan of attack ,
a more general case is treated in the concluding section [ sec : ideaofproof ] , illustrating the use of the methods introduced before .
this paper is dedicated to the memory of eckart viehweg .
like so many other mathematicians of his age group , the author benefited immensely from eckart s presence in the field , his enthusiasm , guidance and support .
eckart will be remembered as an outstanding mathematician , and as a fine human being .
the work on this paper was partially supported by the dfg forschergruppe 790 `` classification of algebraic surfaces and compact complex manifolds '' .
patrick graf kindly read earlier versions of this paper and helped to remove several errors and misprints .
many of the results presented here have been obtained in joint work of sndor kovcs and the author .
the author would like to thank sndor for innumerable discussions , and for a long lasting collaboration .
he would also like to thank the anonymous referee for careful reading and for numerous suggestions that helped to improve the quality of this survey .
not all the material presented here is new , and some parts of this survey have appeared in similar form elsewhere .
the author would like to thank his coauthors for allowing him to use material from their joint research papers .
the first subsection of every chapter lists the sources that the author was aware of .
given its importance in algebraic and arithmetic geometry , much work has been invested to generalise the shafarevich hyperbolicity conjecture , theorem [ thm : shafarevich ] . historically , the first generalisations have been concerned with families @xmath8 where @xmath3 is still a quasi - projective curve , but where the fibres of @xmath7 are allowed to have higher dimension
. the following elementary example shows , however , that theorem [ thm : shafarevich ] can not be generalised navely , and that additional conditions must be posed .
[ ex : counter ] consider a smooth projective surface @xmath0 of general type which contains a rational or elliptic curve @xmath9 .
assume that the automorphism group of @xmath0 fixes the curve @xmath10 pointwise .
examples can be obtained by choosing any surface of general type and then blowing up sufficiently many points in sufficiently general position each blow - up will create a rational curve and lower the number of automorphisms .
thus , if @xmath11 and @xmath12 are any two distinct points , then the surfaces @xmath13 obtained by blowing up the points @xmath14 are non - isomorphic . in order to construct a proper family , consider the product @xmath15 with its projection @xmath16 and with the natural section @xmath17 . if @xmath18 is the blow - up of @xmath15 in @xmath19
, then we obtain a smooth , projective family @xmath20 of surfaces of general type , with the property that no two fibres are isomorphic .
it can well be argued that counterexample [ ex : counter ] is not very natural , and that the fibres of the family @xmath21 would trivially be isomorphic if they had not been blown up artificially .
this might suggest to consider only families that are `` not the blow - up of something else '' .
one way to make this condition is precise is to consider only _ families of minimal surfaces _ , i.e. , surfaces @xmath22 whose canonical bundle @xmath23 is semiample . in higher dimensions
, it is often advantageous to impose a stronger condition and consider only _ families of canonically polarised manifolds _ , i.e. , manifolds @xmath22 whose canonical bundle @xmath23 is ample .
hyperbolicity properties of families of minimal surfaces and families of minimal varieties have been studied by a large number of people , including migliorini @xcite , kovcs @xcite and oguiso - viehweg @xcite . for families of
canonically polarised manifolds , the analogue of theorem [ thm : shafarevich ] has been shown by kovcs in the algebraic setup @xcite . combining algebraic arguments with deep analytic methods , viehweg and zuo
prove a more general brody hyperbolicity theorem for moduli spaces of canonically polarised manifolds which also implies an analogue of theorem [ thm : shafarevich ] , @xcite .
[ thm : kovhyp ] let @xmath1 be a smooth , complex , projective family of canonically polarised varieties of arbitrary dimension , over a smooth quasi - projective base curve @xmath3 .
then the conclusion of the shafarevich hyperbolicity conjecture , theorem [ thm : shafarevich ] , holds .
this paper discusses generalisations of the shafarevich hyperbolicity conjecture to families over higher dimensional base manifolds . to formulate any generalisation
, two points need to be clarified . 1 .
we need to define a higher dimensional analogue for the list of quasi - projective curves given in theorem [ thm : shafarevich ] .
2 . given any family @xmath8 over a higher dimensional base ,
call two points @xmath24 , @xmath25 equivalent if the fibres @xmath26 and @xmath27 are isomorphic .
if @xmath3 is a curve , then either there is only one equivalence class , or all equivalence classes are finite . for families over higher dimensional base manifolds ,
the equivalence classes will generally be subvarieties of arbitrary dimension .
we will need to have a quantitative measure for the number of equivalence classes and their dimensions .
the problems outlined above justify the definition of the _ logarithmic kodaira dimension _ and of the _ variation of a family _ , respectively . before coming to the generalisations of the shafarevich hyperbolicity conjecture in section [ ssec : viehwegconj ] below , we recall the definitions for the reader s convenience .
the logarithmic kodaira dimension generalises the classical notion of kodaira dimension to the category of quasi - projective varieties .
let @xmath3 be a smooth quasi - projective variety and @xmath0 a smooth projective compactification of @xmath3 such that @xmath28 is a divisor with simple normal crossings .
the logarithmic kodaira dimension of @xmath3 , denoted by @xmath29 , is defined to be the kodaira - iitaka dimension of the line bundle @xmath30 .
a quasi - projective variety @xmath3 is called of _ log general type _ if @xmath31 , i.e. , the divisor @xmath32 is big .
it is a standard fact of logarithmic geometry that a compactification @xmath0 with the described properties exists , and that the logarithmic kodaira dimension @xmath29 does not depend on the choice of the compactification .
the quasi - projective curves listed in theorem [ thm : shafarevich ] are precisely those curves @xmath3 with logarithmic kodaira dimension @xmath33 .
the following definition provides a quantitative measure of the _ birational _ variation of a family .
note that the definition is meaningful even in cases where no moduli space exists .
[ def : var ] let @xmath1 be a projective family over an irreducible base @xmath3 , and let @xmath34 denote the algebraic closure of the function field of @xmath3 .
the variation of @xmath7 , denoted by @xmath35 , is defined as the smallest integer @xmath36 for which there exists a subfield @xmath37 of @xmath38 , finitely generated of transcendence degree @xmath36 over @xmath39 and a @xmath37-variety @xmath22 such that @xmath40 is birationally equivalent to @xmath41 . in the setup of definition [ def : var ] , assume that all fibres if @xmath3 are canonically polarised complex manifolds
. then coarse moduli schemes are known to exist , ( * ? ? ?
* thm . 1.11 ) , and the variation is the same as either the dimension of the image of @xmath3 in moduli , or the rank of the kodaira - spencer map at the general point of @xmath3 .
further , one obtains that @xmath42 if and only if all fibres of @xmath7 are isomorphic .
in this case , the family @xmath7 is called `` isotrivial '' . using the notion of `` logarithmic kodaira dimension '' and `` variation '' ,
the shafarevich hyperbolicity conjecture can be reformulated as follows . if @xmath1 is any smooth , complex , projective family of curves of genus @xmath2 , over a smooth quasi - projective base curve @xmath3 , and if @xmath43 , then @xmath44 . aiming to generalise the shafarevich hyperbolicity conjecture to families over higher dimensional base manifolds
, viehweg has conjectured that this reformulation holds true in arbitrary dimension .
[ conj : viehweg ] let @xmath1 be a smooth projective family of varieties with semiample canonical bundle , over a quasi - projective manifold @xmath3 . if @xmath7 has maximal variation , then @xmath3 is of log general type . in other words ,
@xmath45 viehweg s conjecture has been proven by sndor kovcs and the author in case where @xmath3 is a surface , @xcite , or a threefold , @xcite .
the methods developed in these papers will be discussed , and an idea of the proof will be given later in this paper , cf .
the outline of this paper given in section [ ssec : outline ] on page .
[ thm : mainresult0 ] viehweg s conjecture holds in case where @xmath46 . for families of _ canonically polarised _
varieties , much stronger results have been obtained , giving an explicit geometric explanation of theorem [ thm : mainresult0 ] . [
thm : mainresult2 ] let @xmath1 be a smooth projective family of canonically polarised varieties , over a quasi - projective manifold @xmath3 of dimension @xmath46 .
let @xmath0 be a smooth compactification of @xmath3 such that @xmath47 is a divisor with simple normal crossings .
then any run of the minimal model program of the pair @xmath48 will terminate in a kodaira or mori fibre space whose fibration factors the moduli map birationally .
neither the compactification @xmath0 nor the minimal model program discussed in theorem [ thm : mainresult2 ] is unique .
when running the minimal model program , one often needs to choose the extremal ray that is to be contracted . in order to explain the statement of theorem [ thm : mainresult2 ] , let @xmath49 be the appropriate coarse moduli space whose existence is shown , e.g. in ( * ? ?
* thm . 1.11 ) .
further , let @xmath50 be the moduli map associated with the family @xmath7 , and let @xmath51 be the associated rational map from the compactification @xmath0 . if @xmath52 is a rational map obtained by running the minimal model program , and if @xmath53 is the associated kodaira or mori fibre space , then theorem [ thm : mainresult2 ] asserts the existence of a map @xmath54 that makes the following diagram commutative , @xmath55^{\lambda}_{\text{mmp of the pair } ( y , d ) } \ar@{-->}[d]_{\text{moduli map induced by } f^\circ } & & & y_\lambda \ar[d]^{\text{kodaira or mori fibre space } } \\
\mathfrak m & & & z_\lambda \ar@{-->}[lll]^{\exists ! } .}\ ] ] now , if we assume in addition that @xmath56 , then the minimal model program terminates in a kodaira fibre space whose base @xmath57 has dimension @xmath58 , so that @xmath59 .
if we assume that @xmath60 , then the minimal model program terminates in proper mori fibre space and we obtain that @xmath61 and @xmath62 . the following refined answer to viehweg s conjecture
is therefore an immediate corollary of theorem [ thm : mainresult2 ] .
[ cor : mainresult1 ] let @xmath1 be a smooth projective family of canonically polarised varieties , over a quasi - projective manifold @xmath3 of dimension @xmath46
. then either 1 .
@xmath63 and @xmath62 , or 2 .
@xmath56 and @xmath64 .
corollary [ cor : mainresult1 ] asserts that any family of canonically polarised varieties over a base manifold @xmath3 with @xmath65 is necessarily isotrivial .
corollary [ cor : mainresult1 ] has also been shown in case where @xmath3 is a _ projective _ manifold of arbitrary dimension , conditional to the standard conjectures of minimal model theory , cf .
* thm . 1.4 ) .
a very short proof that does not rely on minimal model theory has been announced by patakfalvi as this paper goes to print , @xcite . to see that the result of corollary [ cor : mainresult1 ] is optimal in case @xmath63 , let @xmath66 be any family of canonically polarised varieties with @xmath67 , over a smooth surface @xmath68 ( which may or may not be compact ) . setting @xmath69 and @xmath70
, we obtain a family @xmath71 with variation @xmath72 , and with a base manifold @xmath3 of kodaira dimension @xmath73 . in the setup of corollary [ cor : mainresult1 ]
, if @xmath3 is a projective fano manifold , then a fundamental result of campana and kollr - miyaoka - mori asserts that @xmath3 is rationally connected , ( * ? ? ?
* v. thm .
2.13 ) . in other words ,
given any two points @xmath74 , @xmath75 in @xmath3 , there exists a rational curve @xmath76 which contains both @xmath74 and @xmath75 . recalling from theorem [ thm : kovhyp ] that families over rational curves are isotrivial
, it follows immediately that the family @xmath7 is necessarily isotrivial itself .
a much stronger version of this result has been shown by lohmann , @xcite . given a projective variety @xmath0 and a @xmath77-divisor @xmath78 such that @xmath48 is a divisorially log terminal ( = dlt)we refer to ( * ? ? ?
2.3 ) for the definition of a _ dlt _ pair , and for related notions concerning singularities of pairs that are relevant in minimal model theory . ]
pair , consider the smooth quasi - projective variety @xmath79 lohmann shows that if @xmath48 is log - fano , that is , if the @xmath77-divisor @xmath80 is ample , then any family of canonically polarized varieties over @xmath3 is necessarily isotrivial .
the proof relies on a generalization of araujo s result @xcite which relates extremal rays in the moving cone of a variety with fiber spaces that appear at the end of the minimal model program .
lohmann shows that the moduli map factorizes through any of the fibrations obtained in this way . in a series of papers , including @xcite ,
campana introduced the notion of `` geometric orbifolds '' and `` special varieties '' .
campana s language helps to formulate a very natural generalisation of theorem [ thm : shafarevich ] , which includes the cases covered by the viehweg conjecture [ conj : viehweg ] , and gives ( at least conjecturally ) a satisfactory geometric explanation of isotriviality observed in some families over spaces that are not covered by conjecture [ conj : viehweg ] . before formulating the conjecture ,
we briefly recall the precise definition of a special logarithmic pair for the reader s convenience .
we take the classical bogomolov - sommese vanishing theorem as our starting point .
we refer to @xcite or to the original reference @xcite for an explanation of the sheaf @xmath81 of logarithmic differentials .
[ thm : classbsv ] let @xmath0 be a smooth projective variety and @xmath82 a reduced ( possibly empty ) divisor with simple normal crossings . if @xmath83 is any number and @xmath84 any invertible subsheaf , then the kodaira - iitaka dimension of @xmath85 is at most @xmath86 , i.e. , @xmath87 . in a nutshell
, we say that a pair @xmath88 is special if the inequality in the bogomolov - sommese vanishing theorem is always strict .
[ def : speciallog ] in the setup of theorem [ thm : classbsv ] , a pair @xmath88 is called _ special _ if the strict inequality @xmath89 holds for all @xmath86 and all invertible sheaves @xmath84 . a smooth
, quasi - projective variety @xmath3 is called special if there exists a smooth compactification @xmath0 such that @xmath90 is a divisor with simple normal crossings and such that the pair @xmath48 is special .
it is an elementary fact that if @xmath3 is a smooth , quasi - projective variety and @xmath91 , @xmath92 two smooth compactifications such that @xmath93 are divisors with simple normal crossings , then @xmath94 is special if and only if @xmath95 is .
the notion of special should thus be seen as a property of the quasi - projective variety @xmath3 .
rationally connected manifolds and manifolds @xmath18 with @xmath96 are special . with this notation in place ,
campana s conjecture can be formulated as follows .
[ conj : campana ] let @xmath97 be a smooth family of canonically polarised varieties over a smooth quasi - projective base . if @xmath3 is special , then the family @xmath21 is isotrivial . in analogy with the construction of the maximally rationally connected quotient map of uniruled varieties , campana constructs in (
3 ) an almost - holomorphic `` core map '' whose fibres are special in the sense of definition [ def : speciallog ] .
like the mrc quotient , the core map is uniquely characterised by certain maximality properties , ( * ? ? ?
3.3 ) , which essentially say that the core map of @xmath18 contracts almost all special subvarieties contained in @xmath18 .
if campana s conjecture [ conj : campana ] holds , this would imply that the core map always factors the moduli map , similar to what we have seen in section [ ssec : viehwegconj ] above , @xmath98^{\text{core map}}_{\text{almost holomorphic } } \ar@{-->}[d]_{\text{moduli map induced by } f^\circ } & & & & z \ar@{-->}@/^0.4cm/[dllll]^{\exists ! }
\\ \mathfrak m.}\ ] ] as with viehweg s conjecture [ conj : viehweg ] , campana s conjecture [ conj : campana ] has been shown for surfaces @xcite and threefolds @xcite .
[ thm : main ] campana s conjecture [ conj : campana ] holds if @xmath46 .
viehweg s conjecture [ conj : viehweg ] and campana s conjecture [ conj : campana ] have been shown for families over base manifolds of dimension three or less . as we will see in section [ sec : ideaofproof ] , the restriction to three - dimensional base manifolds comes from the fact that minimal model theory is particularly well - developed for threefolds , and from our limited ability to handle differential forms on singular spaces of higher dimension .
we do not believe that there is a fundamental reason that restricts us to dimension three , and we do believe that the relationship between the moduli map and the mmp found in theorem [ thm : mainresult2 ] will hold in arbitrary dimension . [
conj : mainresult2 ] let @xmath1 be a smooth projective family of canonically polarised varieties , over a quasi - projective manifold @xmath3 .
let @xmath0 be a smooth compactification of @xmath3 such that @xmath99 is a divisor with simple normal crossings .
then any run of the minimal model program of the pair @xmath48 will terminate in a kodaira or mori fibre space whose fibration factors the moduli map birationally .
[ conj : mainresult3 ] corollary [ cor : mainresult1 ] holds without the assumption that @xmath100 .
given the current progress in minimal model theory , a proof of conjectures [ conj : mainresult2 ] and [ conj : mainresult3 ] does no longer seem out of reach .
throughout the present section [ sec : vz ] , we consider a smooth projective family @xmath1 of projective , canonically polarised complex manifolds , over a smooth complex quasi - projective base .
we assume that the family is not isotrivial , and fix a smooth projective compactification @xmath0 of @xmath3 such that @xmath47 is a divisor with simple normal crossings . in this setup ,
viehweg and zuo have shown the following fundamental result asserting the existence of many logarithmic pluri - differentials on @xmath0 .
[ thm : vz ] let @xmath1 be a smooth projective family of canonically polarised complex manifolds , over a smooth complex quasi - projective base .
assume that the family is not isotrivial and fix a smooth projective compactification @xmath0 of @xmath3 such that @xmath101 is a divisor with simple normal crossings .
then there exists a number @xmath102 and an invertible sheaf @xmath103 whose kodaira - iitaka dimension is at least the variation of the family , @xmath104 .
observe that the shafarevich hyperbolicity conjecture , theorem [ thm : shafarevich ] , follows as an immediate corollary of theorem [ thm : vz ] .
a somewhat weaker version of theorem [ thm : vz ] holds for families of projective manifolds with only semiample canonical bundle if one assumes additionally that the family is of maximal variation , i.e. , that @xmath43 , cf .
1.4(iv ) ) .
as we will see in section [ sec : ideaofproof ] , the `` viehweg - zuo '' sheaf @xmath85 is one of the crucial ingredients in the proofs of viehweg s and campana s conjecture for families over threefolds , theorems [ thm : mainresult0 ] , [ thm : mainresult2 ] and [ thm : main ] .
a careful review of viehweg and zuo s construction reveals that the `` viehweg - zuo sheaf '' @xmath85 comes from the coarse moduli space @xmath105 , at least generically .
the precise statement , given in theorem [ thm : vzimproved ] , uses the following notion .
[ not : introb ] let @xmath106 be the moduli map associated with the family @xmath7 , and consider the subsheaf @xmath107 , defined on presheaf level as follows : if @xmath108 is any open set and @xmath109 any section , then @xmath110 if and only if the restriction @xmath111 is in the image of the differential map @xmath112 where @xmath113 is the open subset where the moduli map @xmath114 has maximal rank . by construction
, it is clear that the sheaf @xmath115 is a saturated subsheaf of @xmath116 , i.e. , that the quotient sheaf @xmath117 is torsion free .
we say that @xmath115 is the saturation of @xmath118 in @xmath119 .
[ thm : vzimproved ] in the setup of theorem [ thm : vz ] , there exists a number @xmath102 and an invertible subsheaf @xmath120 whose kodaira - iitaka dimension is at least the variation of the family , @xmath121 .
theorem [ thm : vzimproved ] follows without too much work from viehweg s and zuo s original arguments and constructions , which are reviewed in section [ ssec : vz2 ] below . compared with theorem [ thm : vz ]
, the refined viehweg - zuo theorem relates more directly to campana s conjecture [ conj : campana ] and other generalizations of the shafarevich conjecture . to illustrate its use , we show in the surface case how theorem [ thm : vzimproved ] reduces campana s conjecture [ conj : campana ] to the viehweg conjecture [ conj : viehweg ] , for which a positive answer is known .
conjecture [ conj : campana ] holds if @xmath122 .
we maintain the notation of conjecture [ conj : campana ] and let @xmath123 be a smooth family of canonically polarised varieties over a smooth quasi - projective base , with @xmath3 a special surface .
since @xmath3 is special , it is not of log general type , and the solution to viehweg s conjecture in dimension two , ( * ? ? ?
1.1 ) , gives that @xmath124 .
we argue by contradiction , suppose that @xmath125 and choose a compactification @xmath88 as in definition [ def : speciallog ] .
by theorem [ thm : vzimproved ] there exists a number @xmath126 and an invertible subsheaf @xmath127 such that @xmath128 .
however , since @xmath115 is saturated in the locally free sheaf @xmath129 , it is reflexive , ( * ? ? ?
* claim on p. 158 ) , and since @xmath125 , the sheaf @xmath115 is of rank @xmath130 .
thus @xmath131 is an invertible subsheaf , ( * ? ? ?
* lem . 1.1.15 , on p. 154 ) , and definition [ def : speciallog ] of a special pair gives that @xmath132 , contradicting the fact that @xmath133 .
it follows that @xmath42 and that the family is hence isotrivial .
given its importance in the theory , we give a very brief synopsis of viehweg - zuo s proof of theorem [ thm : vz ] , showing how the theorem follows from deep positivity results and fact [ fact : vz48 ] , respectively . ] for push - forward sheaves of relative dualizing sheaves , and for kernels of kodaira - spencer maps , respectively . even though no proof of the refined theorem [ thm : vzimproved ] , is given
, it is hoped that the reader who chooses to read section [ ssec : vz2 ] will believe that theorem [ thm : vzimproved ] follows with some extra work by essentially the same methods .
the reader who is interested in a detailed understanding , including is referred to the papers @xcite , @xcite , and to the survey @xcite .
the overview contained in this section and the facts outlined in section [ ssec : facts ] can perhaps serve as a guideline to @xcite .
many of the technical difficulties encountered in the full proof of theorem [ thm : vz ] vanish if @xmath7 is a family of curves .
the proof becomes very transparent in this case .
in particular , it is very easy to see how the kodaira - spencer map associated with the family @xmath7 transports the positivity found in push - forward sheaves into the sheaf of differentials @xmath134 .
after setting up notation in section [ ssec:8a ] , we have therefore included a section [ ssec:8b ] which discusses the curve case in detail .
most of the material presented in the current section [ sec : vz ] , including the synopsis of viehweg - zuo s construction , is taken without much modification from the paper @xcite .
the presentation is inspired in part by @xcite . throughout the present section [ ssec : vz2 ] ,
we choose a smooth projective compactification @xmath18 of @xmath135 such that the following holds : 1 .
the difference @xmath136 is a divisor with simple normal crossings .
the morphism @xmath7 extends to a projective morphism @xmath137 .
it is then clear that @xmath138 set - theoretically . removing a suitable subset @xmath139 of codimension @xmath140 , the following
will then hold automatically on @xmath141 and @xmath142 , respectively . 1 .
the restricted morphism @xmath143 is flat .
the divisor @xmath144 is smooth .
the divisor @xmath145 is a relative normal crossing divisor , i.e. a normal crossing divisor whose components and all their intersections are smooth over the components of @xmath146 . in the language of viehweg - zuo , (
* def 2.1(c ) ) , the restricted morphism @xmath147 is a `` good partial compactification of @xmath7 '' . [
rem : extos ] let @xmath148 be a locally free sheaf on @xmath0 , and let @xmath149 be an invertible subsheaf . since @xmath150
, there exists a unique extension of the sheaf @xmath151 to an invertible subsheaf @xmath152 on @xmath0 .
furthermore , the restriction map @xmath153 is an isomorphism . in particular , the notion of kodaira - iitaka dimension makes sense for the sheaf @xmath151 , and @xmath154 .
we denote the relative dimension of @xmath18 over @xmath0 by @xmath155 . before sketching the proof of theorem [ thm : vz ] in full generality
, we illustrate the main idea in a particularly simple setting .
[ sass : famcurves ] throughout the present introductory section [ ssec:8b ] , we maintain the following simplifying assumptions in addition to the assumptions made in theorem [ thm : vz ] . 1 .
the quasi - projective variety @xmath3 is in fact projective .
in particular , we assume that @xmath156 , @xmath157 , that @xmath158 and that @xmath159 .
2 . the family @xmath137 is a family of curves of genus @xmath2 .
in particular , we have that @xmath160 , where @xmath161 is the kernel of the derivative @xmath162 .
the variation of @xmath7 is maximal , that is , @xmath163 .
the proof of theorem [ thm : vz ] sketched here uses positivity of the push - forward of relative dualizing sheaves as its main input .
the positivity result required is discussed in viehweg s survey ( * ? ? ?
13 ) , where positivity is obtained as a consequence of generalised kodaira vanishing theorems .
the reader interested in a broader overview might also want to look at the remarks and references in ( * ? ? ?
6.3.e ) , as well as the papers @xcite [ thm : vpos ] under the simplifying assumptions [ sass : famcurves ] , the push - forward sheaf @xmath164 is locally free of positive rank .
if @xmath165 is any ample line bundle , then there exist numbers @xmath166 and a sheaf morphism @xmath167 which is surjective at the general point of @xmath0 .
for the reader s convenience , we recall two other facts used in the proof , namely the existence of a kodaira - spencer map , and serre duality in the relative setting .
[ thm : kodaira - spencer ] under the simplifying assumptions [ sass : famcurves ] , since @xmath168 , there exists a non - zero sheaf morphism @xmath169 which measures the variation of the isomorphism classes of fibres in moduli .
[ thm : serre ] under the simplifying assumptions [ sass : famcurves ] , if @xmath170 is any coherent sheaf on @xmath18 , then there exists a natural isomorphism @xmath171 .
consider the dual of the ( non - trivial ) kodaira - spencer map discussed in theorem [ thm : kodaira - spencer ] , say @xmath172 .
recalling that @xmath173 equals the relative dualizing sheaf @xmath174 , and using the relative serre duality theorem [ thm : serre ] , the sheaf morphism @xmath175 is naturally identified with a non - zero morphism @xmath176 choosing an ample line bundle @xmath165 and sufficiently large and divisible numbers @xmath166 , theorem [ thm : vpos ] yields a sequence of sheaf morphisms @xmath177^(.4){\phi}_(.4){\text{gen.~surjective } } & & \operatorname{sym}^n f _ * \bigl(\omega_{x / y}^{\otimes 2 } \bigr ) \ar[rr]^(.55){\operatorname{sym}^n(\kappa^*)}_(.55){\text{non - trivial } } & & \operatorname{sym}^n \omega^1_y , } \ ] ] whose composition @xmath178 is clearly not the zero map .
consequently , we obtain a non - trivial map @xmath179 , finishing the proof of theorem [ thm : vz ] under the simplifying assumptions [ sass : famcurves ] .
the proof outlined above uses the dual of the kodaira - spencer map as a vehicle to transport the positivity which exists in @xmath180 into the sheaf @xmath181 of differential forms on @xmath0 .
if @xmath21 was a family of surfaces rather than a family of curves , then serre duality can not easily be used to identify the dual of @xmath182 with a push - forward sheaf of type @xmath183 , or any with other sheaf whose positivity is well - known . to overcome this problem
, viehweg suggested to replace the kodaira - spencer map @xmath184 by sequences of more complicated sheaf morphisms @xmath185 and @xmath186 , constructed in sections [ sssec : hks1 ] and [ sssec : hks2 ] below . to motivate the slightly involved construction of these maps , we recall without proof a description of the classical kodaira - spencer map . under the simplifying assumptions [ sass : famcurves ] , consider the standard sequence of relative differential forms on @xmath18 , @xmath187 and its twist with the invertible sheaf @xmath188 , @xmath189 using that @xmath190 , the first connecting morphism associated with this sequence then reads @xmath191 the sheaf @xmath170 is naturally isomorphic to the sheaf hom@xmath192 . to give a morphism @xmath193
is thus the same as to give a map @xmath194 , and the morphism obtained in is the same as the kodaira - spencer map discussed in theorem [ thm : kodaira - spencer ] .
observe also that serre duality yields a natural identification of @xmath195 with the sheaf hom@xmath196 . to give a morphism @xmath193 it is thus the same as to give a map @xmath197 .
the morphism obtained in this way from is of course the morphism @xmath175 of equation . in the general
setting of theorem [ thm : vz ] where the simplifying assumptions [ sass : famcurves ] do not generally hold , the starting point of the viehweg - zuo construction is the standard sequence of relative logarithmic differentials associated to the flat morphism @xmath199 which generalises sequence from above , @xmath200 we refer to ( * ? ? ?
4 ) for a discussion of sequence , and for a proof of the fact that the cokernel @xmath201 is locally free . by (
* ii , ex . 5.16 ) , sequence defines a filtration of the @xmath202 exterior power , @xmath203 with @xmath204 .
take the first sequence induced by the filtration , @xmath205 modulo @xmath206 , and obtain @xmath207 setting @xmath208 , twisting sequence with @xmath209 and pushing down , the connecting morphisms of the associated long exact sequence give maps @xmath210 where @xmath211 .
set @xmath212 .
the morphisms @xmath185 and @xmath213 can be composed if we tensor the latter with the identity morphism on @xmath214 .
more specifically , we consider the following morphisms , @xmath215 and their compositions @xmath216 theorem [ thm : vz ] is shown by relating the morphism @xmath185 with the structure morphism of a higgs - bundle coming from the variation of hodge structures associated with the family @xmath7 .
viehweg s positivity results of push - forward sheaves of relative differentials , as well as zuo s results on the curvature of kernels of generalised kodaira - spencer maps are the main input here . rather than recalling the complicated line of argumentation ,
we simply state two central results from the argumentation of @xcite .
[ fact : vz46 ] for any @xmath219 , the morphism @xmath186 factors via the symmetric differentials @xmath220 .
more precisely , the morphism @xmath186 takes its image in @xmath221 .
using fact [ fact : vz46 ] and the observation that @xmath222 , we can therefore view @xmath186 as a morphism @xmath223 while the proof of fact [ fact : vz46 ] is rather elementary , the following deep result is at the core of viehweg - zuo s argument .
its role in the proof of theorem [ thm : vz ] is comparable to that of the positivity theorem [ thm : vpos ] discussed in section [ ssec:8b ] .
[ fact : vz48 ] given any numbers @xmath86 and @xmath224 , there exists a number @xmath219 and an invertible sheaf @xmath225 of kodaira - iitaka dimension @xmath226 such that @xmath227 is generically generated . to end the sketch of proof , we follow @xcite almost verbatim . by fact
[ fact : vz48 ] , the trivial sheaf @xmath228 can not lie in the kernel @xmath229 of @xmath230 .
we can therefore set @xmath231 to be the largest number with @xmath232 . since @xmath233 is maximal , @xmath234 and @xmath235 in other words
, @xmath236 gives a non - trivial map @xmath237 equivalently , we can view @xmath236 as a non - trivial map @xmath238 by fact [ fact : vz48 ] , there are many morphisms @xmath239 , for @xmath219 large enough .
together with , this gives a non - zero morphism @xmath240 .
we have seen in remark [ rem : extos ] that the sheaf @xmath241 extends to a sheaf @xmath242 with @xmath243 .
this ends the proof of theorem [ thm : vz ] . in spite of its importance ,
little is known about further properties that the viehweg - zuo sheaves @xmath244 might have .
for families of higher - dimensional manifolds , how does the viehweg - zuo construction behave under base change ?
does it satisfy any universal properties at all ?
if not , is there a `` natural '' positivity result for base spaces of families that does satisfy good functorial properties ? in the setup of theorem [ thm : vz ] , if @xmath245 is any closed submanifold , then the associated viehweg - zuo sheaves @xmath244 , constructed for the family @xmath8 , and @xmath246 , constructed for the restricted family @xmath247 , may differ .
in particular , it is not clear that @xmath248 is the restriction of @xmath244 , and the sheaves @xmath244 and @xmath248 may live in different symmetric products of their respective @xmath249 s .
one likely source of non - compatibility with base change is the choice of the number @xmath233 in section [ ssec : eopvz ] ( `` largest number with @xmath232 '' ) .
it seems unlikely that this definition behaves well under base change . for families of higher - dimensional manifolds
, are there distinguished subvarieties in moduli space that have special viehweg - zuo sheaves , perhaps contained in particularly high / low symmetric powers of @xmath249 ?
does the lack of a restriction morphism induce a geometric structure on the moduli space ?
the refinement of the viehweg - zuo theorem , presented in theorem [ thm : vzimproved ] above , turns out to be important for the applications that we have in mind .
it is , however , not clear to us if the sheaf @xmath115 which appears in theorem [ thm : vzimproved ] is really optimal . prove that the sheaf @xmath250 is the smallest sheaf possible for which theorem [ thm : vzimproved ] holds , or else find the smallest possible sheaf .
for instance , does theorem [ thm : vzimproved ] admit a natural improvement if we replace @xmath119 by a suitable sheaf of orbifold differentials , using campana s language of geometric orbifolds ?
to motivate the results presented in this section , let @xmath251 be a smooth , projective family of canonically polarised varieties over a smooth , quasi - projective base manifold , and assume that the family @xmath7 is of maximal variation , i.e. , that @xmath252 .
as before , choose a smooth compactification @xmath253 such that @xmath47 is a divisor with only simple normal crossings . to prove viehweg s conjecture
, we need to show that the logarithmic kodaira dimension of @xmath3 is maximal , i.e. , that @xmath254 .
in particular , we need to rule out the possibility that @xmath255 .
as we will see in the proof of proposition [ prop : vzec1 ] below , a relatively elementary argument exists in cases where the picard number of @xmath0 is one , @xmath256 .
we refer the reader to ( * ? ? ?
i.1 ) for the notion of semistability and for a discussion of the harder - narasimhan filtration used in the proof .
[ prop : vzec1 ] in the setup described above , if we additionally assume that @xmath256 , then @xmath257 .
we argue by contradiction and assume that both @xmath258 and that @xmath256 . let @xmath259 be the big invertible sheaf whose existence is guaranteed by the viehweg - zuo construction , theorem [ thm : vz ] . since @xmath256 ,
the sheaf @xmath85 is actually ample . as a first step ,
observe that the log canonical bundle @xmath32 must be torsion , i.e. , that there exists a number @xmath260 such that @xmath261 .
this follows from the assumption that @xmath262 and from the observation that on a projective manifold with @xmath263 , any invertible sheaf which admits a non - zero section is either trivial or ample .
in particular , we obtain that the divisor @xmath32 is numerically trivial . next , let @xmath264 be a general complete intersection curve in the sense of mehta - ramanathan , cf .
ii.7 ) . the numerical triviality of @xmath32
will then imply that @xmath265 on the other hand , since @xmath85 is ample , we have that @xmath266 . in summary
, we obtain that the symmetric product sheaf @xmath267 is not semistable .
since we are working in characteristic zero , this implies that the sheaf of khler differentials @xmath119 will likewise not be semistable , and contains a destabilising subsheaf @xmath250 with @xmath268 , cf .
since the intersection number @xmath269 is positive , the rank @xmath270 of the sheaf @xmath115 must be strictly less than @xmath271 , and its determinant is an ample invertible subsheaf of the sheaf of logarithmic @xmath270-forms , @xmath272 this , however , contradicts the bogomolov - sommese vanishing theorem [ thm : classbsv ] and therefore ends the proof .
the assumption that @xmath256 is not realistic . in the general situation , where @xmath273 can be arbitrarily large
, we will apply the minimal model program to the pair @xmath48 . as we will sketch in section [ sec : ideaofproof ] , assuming that the standard conjectures of minimal model theory hold true , a run of the minimal model program for a suitable choice of a boundary divisor will yield a diagram , @xmath274^{\lambda}_{\text{minimal model program } } & & & y_\lambda \ar[d]^{\text{fibre space}}_{\pi } \\ & & & z_\lambda , } \ ] ] where @xmath52 is a birational map whose inverse does not contract any divisors , and where either @xmath275 and @xmath57 is a point , or where @xmath276 has the structure of a proper mori or kodaira fibre space . in the first case
, we can try to copy the proof of proposition [ prop : vzec1 ] above . in the second case , we can use the fibre structure and try to argue inductively . the main problem that arises
when adopting the proof of proposition [ prop : vzec1 ] is the presence of singularities .
both the space @xmath276 and the cycle - theoretic image divisor @xmath277 will generally be singular , and the pair @xmath278 will generally be dlt .
this leads to two difficulties . 1 .
the sheaf @xmath279 of logarithmic khler differentials is generally not pure in the sense of ( * ? ? ?
accordingly , there is no good notion of stability that would be suitable to construct a harder - narasimhan filtration .
2 . [ x ]
the viehweg - zuo construction does not work for singular varieties .
the author is not aware of any method suitable to prove positivity results for khler differentials , or prove the existence of sections in any symmetric product of @xmath279 .
the aim of the present section [ sec : ext ] is to show that that both problems can be overcome if we replace the sheaf @xmath280 of khler differentials by its double dual @xmath281}_{y_\lambda}(\log d_\lambda ) : = \bigl ( \omega^1_{y_\lambda}(\log d_\lambda ) \bigr)^{**}$ ] .
we refer to ( * ? ? ?
1.6 ) for a discussion of the double dual in this context , and to ( * ? ? ?
1.1 ) for a thorough discussion of reflexive sheaves
. the following notation will be useful in the discussion .
[ not : relfxive ] let @xmath18 be a normal variety and @xmath244 a coherent sheaf of @xmath282-modules .
given any number @xmath283 , set @xmath284 } : = ( \mathcal a^{\otimes n})^{**}$ ] , @xmath285 } \mathcal a : = ( \operatorname{sym}^n \mathcal a)^{**}$ ] .
if @xmath286 is a morphism of normal varieties , set @xmath287}(\mathcal a ) : = \bigl ( \pi^*\mathcal a \bigr)^{**}$ ] . in a similar vein ,
set @xmath288}_x : = \bigl ( \omega^p_x \bigr)^{**}$ ] and @xmath288}_x(\log d ) : = \bigl ( \omega^p_x ( \log d ) \bigr)^{**}$ ] . a section in @xmath288}_x$ ] or @xmath288}_x(\log d)$ ] will be called a _ reflexive form _ or a _ reflexive logarithmic form _ , respectively .
reflexive sheaves are torsion free and therefore pure . in particular , a harder - narasimhan filtration exists for @xmath288}_x(\log d)$ ] and for the symmetric products @xmath285 } \omega^1_x(\log d)$ ] .
[ fact : extension1 ] if @xmath18 is a normal space , if @xmath244 is any reflexive sheaf on @xmath18 and if @xmath289 any set of @xmath290 , then the restriction map @xmath291 is in fact isomorphic .
we say that `` sections in @xmath244 extend over the small set @xmath292 '' .
if @xmath293 is the complement of @xmath292 , with inclusion map @xmath294 , it follows immediately that @xmath295 . in a similar vein ,
if @xmath296 is any locally free sheaf on @xmath297 , its push - forward sheaf @xmath298 will always be reflexive .
it follows almost by definition that sheaves of reflexive differentials have very good push - forward properties . in section [ ssec : pfwd ] we will use these properties to overcome one of the problems mentioned above and to produce viehweg - zuo sheaves of reflexive differentials on singular spaces . perhaps more importantly , we will in section [ ssec : pb ] recall extension results for log canonical varieties .
these results show that reflexive differentials often admit a pull - back map , similar to the standard pull - back of khler differentials .
a generalisation of the bogomolov - sommese vanishing theorem to log canonical varieties follows as a corollary . in section [ ssec : res ]
, we recall that some of the most important constructions known for logarithmic differentials on snc pairs also work for reflexive differentials on dlt pairs .
this includes the existence of a residue sequence . for our purposes , this makes reflexive differentials almost as useful as regular differentials in the theory of smooth spaces . as we will roughly sketch in section [ sec : ideaofproof ] , these results will allow to adapt the proof of proposition [ prop : vzec1 ] to the singular setup , and will give a proof of viehweg s conjecture [ conj : viehweg ] , at least for families over base manifolds of dimension @xmath299 .
section [ ssec : ioppb ] gives a brief sketch of the proof of the pull - back result of section [ ssec : pb ] .
we end by mentioning a few open problems and conjectures .
some of the material presented in the current section [ sec : ext ] , including section [ ssec : res ] and all the illustrations , is taken without much modification from the paper @xcite .
section [ ssec : pfwd ] follows the paper @xcite .
fact [ fact : extension1 ] implies that any viehweg - zuo sheaf which exists on a pair @xmath300 of a smooth variety and a reduced divisor with simple normal crossing support immediately implies the existence of a viehweg - zuo sheaf of reflexive differentials on any minimal model of @xmath300 , and that the kodaira - iitaka dimension only increases in the process . to formulate the result
precisely , we briefly recall the definition of the kodaira - iitaka dimension for reflexive sheaves .
[ def : kidim ] let @xmath292 be a normal projective variety and @xmath85 a reflexive sheaf of rank one on @xmath292 . if @xmath301}\bigr ) = 0 $ ] for all @xmath302 , then we say that @xmath85 has kodaira - iitaka dimension @xmath303 . otherwise , set @xmath304}\bigr)>0\bigr\},\ ] ] recall that the restriction of @xmath85 to the smooth locus of @xmath292 is locally free and consider the natural rational mapping @xmath305}\bigr)^*\bigr ) \text { for each } n \in m.\ ] ] the kodaira - iitaka dimension of @xmath85 is then defined as @xmath306 with this notation , the main result concerning the push - forward is then formulated as follows .
[ lem : pushdowna ] let @xmath300 be a pair of a smooth variety and a reduced divisor with simple normal crossing support . assume that there exists a reflexive sheaf @xmath307 } \omega^1_z(\log \delta)$ ] of rank one .
if @xmath308 is a birational map whose inverse does not contract any divisor , if @xmath309 is normal and @xmath310 is the ( necessarily reduced ) cycle - theoretic image of @xmath19 , then there exists a reflexive sheaf @xmath311 } \omega^1_{z'}(\log \delta')$ ] of rank one , and of kodaira - iitaka dimension @xmath312 .
the assumption that @xmath313 does not contract any divisors and the normality of @xmath309 guarantee that @xmath314 is a well - defined embedding over an open subset @xmath315 whose complement has codimension @xmath316 , cf .
zariski s main theorem ( * ? ? ?
* v 5.2 ) .
in particular , @xmath317 .
let @xmath318 denote the inclusion and set @xmath319 this sheaf is reflexive by fact [ fact : extension1 ] .
we obtain an inclusion of reflexive sheaves , @xmath320 } \omega^1_{z'}(\log \delta')$ ] . by construction
, we have that @xmath321}\bigr ) \geq h^0(z,\ , \operatorname{\mathcal a}^{[m]})$ ] for all @xmath126 , hence @xmath322 . given the importance of the viehweg - zuo construction , theorem [ thm : vz ] , we will call the sheaves @xmath244 which appear in proposition [ lem : pushdowna ] `` viehweg - zuo sheaves '' .
let @xmath300 be a pair of a smooth variety and a reduced divisor with simple normal crossing support , and let @xmath323 be any number .
a reflexive sheaf @xmath307 } \omega^1_z(\log \delta)$ ] of rank one will be called a `` viehweg - zuo sheaf '' .
khler differentials are characterised by a number of universal properties , one of the most important being the existence of a pull - back map : if @xmath324 is any morphism of algebraic varieties and if @xmath325 , then there exists a canonically defined sheaf morphism @xmath326 the following example illustrates that for sheaves of reflexive differentials on normal spaces , a pull - back map does not exist in general .
let @xmath18 be a normal gorenstein variety of dimension @xmath327 , and let @xmath328 be any resolution of singularities .
observing that the sheaf of reflexive @xmath327-forms is precisely the dualizing sheaf , @xmath329}_x \simeq \omega_x$ ] , it follows directly from the definition of canonical singularities that @xmath18 has canonical singularities if and only if a pull - back morphism @xmath330}_x \to \omega^{n}_z$ ] exists .
together with daniel greb , sndor kovcs and thomas peternell , the author has shown that a pull - back map for reflexive differentials always exists if the target is log canonical .
[ thm : generalpullback ] let @xmath331 be an log canonical pair , and let @xmath332 be a morphism from a normal variety @xmath292 such that the image of @xmath292 is not contained in the reduced boundary or in the singular locus , i.e. , @xmath333 if @xmath334 is any index and @xmath335 then there exists a sheaf morphism , @xmath336}_x(\log \lfloor d \rfloor ) \to \omega^{[p]}_z(\log \delta),\ ] ] that agrees with the usual pull - back morphism of khler differentials at all points @xmath337 where @xmath338 . if follows from the definition of klt , ( * ? ? ?
* def . 2.34 ) , that the components of @xmath78 which appear with coefficient one are always contained in the non - klt locus of @xmath339 .
in particular , the divisor @xmath19 defined in theorem [ thm : generalpullback ] always contains the codimension - one part of @xmath340 .
the assertion of theorem [ thm : generalpullback ] is rather general and perhaps a bit involved . for klt spaces
, the statement reduces to the following simpler result .
let @xmath18 be a normal klt variety be a normal variety such that the pair @xmath341 is klt '' ] , and let @xmath332 be a morphism from a normal variety @xmath292 such that the image @xmath342 is not entirely contained in the singular locus of @xmath18 . if @xmath334 is any index then there exists a sheaf morphism , @xmath336}_x \to \omega^{[p]}_z,\ ] ] that agrees on an open set with the usual pull - back morphism of khler differentials .
extension properties of differential forms that are closely related to the existence of pull - back maps have been studied in the literature , mostly considering only special values of @xmath86 . using steenbrink s generalization of the grauert - riemenschneider vanishing theorem as their main input ,
similar results were shown by steenbrink and van straten for varieties @xmath18 with only isolated singularities and for @xmath343 , without any further assumption on the nature of the singularities , ( * ? ? ?
flenner extended these results to normal varieties , subject to the condition that @xmath344 , @xcite .
namikawa proved the extension properties for @xmath345 , in case @xmath18 has canonical gorenstein singularities , ( * ? ? ?
* thm . 4 ) .
in the case of finite quotient singularities similar results were obtained in @xcite . for a log canonical pair with reduced boundary divisor , the cases @xmath346 were settled in ( * ? ? ?
1.1 ) . a related setup where the pair @xmath339 is snc , and where @xmath347 is the composition of a finite galois covering and a subsequent resolution of singularities has been studied by esnault and viehweg . in @xcite
they obtain in their special setting similar results and additionally prove vanishing of higher direct image sheaves .
a brief sketch of the proof of theorem [ thm : generalpullback ] is given in section [ ssec : ioppb ] below .
the proof uses a strengthening of the steenbrink vanishing theorem , which follows from local hodge - theoretic properties of log canonical singularities , in particular from the fact that log canonical spaces are du bois .
these methods are combined with results available only for special classes of singularities , such as the recent progress in minimal model theory and partial classification of singularities that appear in minimal models .
theorem [ thm : generalpullback ] has many applications useful for moduli theory .
we mention two applications which will be important in our context .
the first application generalises the bogomolov - sommese vanishing theorem to singular spaces .
[ cor : bs ] let @xmath331 be a log canonical pair , where @xmath18 is projective .
if @xmath348}_x(\log \lfloor d \rfloor)$ ] is a @xmath77-cartier reflexive subsheaf of rank one , then @xmath349 .
the number @xmath350 appearing in the statement of corollary [ cor : bs ] is the generalised kodaira - iitaka dimension introduced in definition [ def : kidim ] .
a reflexive sheaf @xmath85 is rank one is called @xmath77-cartier if there exists a number @xmath351 such that the @xmath327th reflexive tensor product @xmath352}$ ] is invertible .
we prove corollary [ cor : bs ] only in the special case where the sheaf @xmath348}_x(\log \lfloor d \rfloor)$ ] is invertible .
the reader interested in a full proof is referred to the original reference @xcite .
let @xmath353 be any resolution of singularities , and let @xmath354 be the reduced divisor defined in theorem [ thm : generalpullback ] above .
theorem [ thm : generalpullback ] will then assert the existence of an inclusion @xmath355 and the standard bogomolov - sommese vanishing result , theorem [ thm : classbsv ] , applies to give that @xmath356 . since @xmath85 is invertible , and
since @xmath357 is birational , it is clear that @xmath358 , finishing the proof .
taking the double dual of a sheaf does generally _ not _ commute with pulling back .
since reflexive tensor products were used in definition [ def : kidim ] to define the kodaira - iitaka dimension of a sheaf , it is generally false that the kodaira - iitaka dimension stays invariant when pulling a sheaf @xmath85 back to a resolution of singularities .
the proof of corollary [ cor : bs ] which is given in the simple case where @xmath85 is invertible does therefore not work without substantial modification in the general setup where @xmath85 is only @xmath359-cartier .
the second application of theorem [ thm : generalpullback ] concerns rationally chain connected singular spaces .
rationally chain connected manifolds are rationally connected , and do not carry differential forms . building on work of hacon and mckernan , @xcite , we show that the same holds for reflexive forms on klt pairs .
[ cor : kltrcc ]
let @xmath339 be a klt pair .
if @xmath18 is rationally chain connected , then @xmath18 is rationally connected , and @xmath360}_x \bigr ) = 0 $ ] for all @xmath361 . choose a log resolution of singularities , @xmath362 of the pair @xmath339 .
since klt pairs are also dlt , a theorem of hacon - mckernan , ( * ? ? ?
1.5(2 ) ) , applies to show that @xmath18 and @xmath363 are both rationally connected . in particular
, it follows that @xmath364 for all @xmath365 by ( * ? ? ?
since @xmath339 is klt , theorem [ thm : generalpullback ] asserts that there exists a pull - back morphism @xmath366}_x \to \omega^p_{\widetilde{x}}$ ] .
as @xmath367 is birational , @xmath368 is generically injective and since @xmath288}_x$ ] is torsion - free , this means that the induced morphism on the level of sections is injective : @xmath369}_x \bigr ) \to h^0 \bigl ( \widetilde{x},\ , \omega^p_{\widetilde{x } } \bigr ) = 0.\ ] ] the claim then follows .
logarithmic khler differentials on snc pairs are canonically defined .
they are characterised by strong universal properties and appear accordingly in a number of important sequences , filtered complexes and other constructions .
first examples include the following : 1 . the pull - back property of differentials under arbitrary morphisms , 2 .
relative differential sequences for smooth morphisms , 3 .
residue sequences associated with snc pairs , and 4 . the description of chern classes as the extension classes of the first residue sequence .
reflexive differentials do in general not enjoy the same universal properties as khler differentials .
however , we have seen in theorem [ thm : generalpullback ] that reflexive differentials do have good pull - back properties if we are working with log canonical pairs . in the present section [ ssec : res ] , we would like to make the point that each of the other properties listed above also has a very good analogue for reflexive differentials , as long as we are working with dlt pairs .
this makes reflexive differential extremely useful in practise . in a sense
, it seems fair to say that `` reflexive differentials and dlt pairs are made for one another '' .
here we recall the generalisation of the standard sequence for relative differentials , ( * ? ? ?
ii.8.11 ) , to the logarithmic setup . for this , we introduce the notion of an _
snc morphism _ as a logarithmic analogue of a smooth morphism .
although _ relatively snc divisors _ have long been used in the literature , cf .
3 ) , we are not aware of a good reference that discusses them in detail . recall that a pair @xmath339 is called an `` snc pair '' if @xmath18 is smooth , and if the divisor @xmath78 is reduced and has only simple normal crossing support . [ not : di ] let @xmath339 be a pair of a normal space @xmath18 and a divisor @xmath78 , where @xmath78 is written as a sum of its irreducible components @xmath370 .
if @xmath371 is any non - empty subset , we consider the scheme - theoretic intersection @xmath372 . if @xmath373 is empty , set @xmath374 .
[ rem : descrsnc ] in the setup of notation [ not : di ] , it is clear that the pair @xmath339 is snc if and only if all @xmath375 are smooth and of codimension equal to the number of defining equations : @xmath376 for all @xmath373 where @xmath377 .
[ def : sncmorphism ] if @xmath339 is an snc pair and @xmath378 a surjective morphism to a smooth variety , we say that @xmath78 is _ relatively snc _ , or that @xmath379 is _ an snc morphism of the pair _
@xmath339 if for any set @xmath373 with @xmath377 all restricted morphisms @xmath380 are smooth of relative dimension @xmath381 .
[ rem : fibersnc ] if @xmath339 is an snc pair and @xmath378 is any surjective snc morphism of @xmath339 , it is clear from remark [ rem : descrsnc ] that if @xmath382 is any point , with preimages @xmath383 and @xmath384 then the pair @xmath385 is again snc .
[ rem : gensnc ] if @xmath339 is an snc pair and @xmath378 is any surjective morphism , it is clear from generic smoothness that there exists a dense open set @xmath386 , such that @xmath387 is relatively snc over @xmath388 .
let @xmath339 be a reduced snc pair , and @xmath378 an snc morphism of @xmath339 , as introduced in definition [ def : sncmorphism ] . in this setting
, the standard pull - back morphism of 1-forms extends to yield the following exact sequence of locally free sheaves on @xmath18 , @xmath389 called the `` relative differential sequence for logarithmic differentials '' . we refer to ( * ? ? ?
4.1 ) ( * ? ? ?
3.3 ) or @xcite for a more detailed explanation . for forms of higher degrees ,
the sequence induces filtration @xmath390 with quotients @xmath391 for all @xmath270 .
we refer to ( * ? ? ?
ii.5.16 ) for the construction of .
the main result of this section , theorem [ thm : relativedifferentialfiltration ] , gives analogues of in case where @xmath339 is dlt .
in essence , theorem [ thm : relativedifferentialfiltration ] says that all properties of the relative differential sequence still hold on dlt pairs if one removes from @xmath18 a set @xmath292 of codimension @xmath392 .
[ thm : relativedifferentialfiltration ] let @xmath331 be a dlt pair with @xmath18 connected .
let @xmath378 be a surjective morphism to a normal variety @xmath393 .
then , there exists a non - empty smooth open subset @xmath394 with preimages @xmath395 , @xmath396 , and a filtration @xmath397}_{x^\circ}(\log \lfloor d^{\circ } \rfloor ) = \mathcal f^{[0]}(\log ) \supseteq \dots \supseteq \mathcal f^{[p]}(\log)\supseteq \{0\}\ ] ] on @xmath135 with the following properties . 1 . [ il:_a ]
the filtrations and agree wherever the pair @xmath398 is snc , and @xmath379 is an snc morphism of @xmath399 .
[ il:_b ] for any @xmath270 , the sheaf @xmath400}(\log)$ ] is reflexive , and @xmath401}(\log)$ ] is a saturated subsheaf of @xmath400}(\log)$ ] .
[ seq : quotsf ] for any @xmath270 , there exists a sequence of sheaves of @xmath402-modules , @xmath403}(\log ) \to \mathcal f^{[r]}(\log ) \to
\omega_{x^\circ / t^\circ}^{[p - r]}(\log \lfloor d^\circ \rfloor ) \to 0,\ ] ] which is exact and analytically locally split in codimension @xmath404 .
[ lastisom ] there exists an isomorphism @xmath405}(\log ) \simeq
\phi^*\omega^p_{t^\circ}$ ] .
if @xmath406 is any complex variety , we call a sequence of sheaf morphisms , @xmath407 `` exact and analytically locally split in codimension @xmath404 '' if there exists a closed subvariety @xmath408 of codimension @xmath409 and a covering of @xmath410 by subsets @xmath411 which are open in the analytic topology , such that the restriction of to @xmath412 is exact , and such that the restriction of to any of the open sets @xmath413 splits . we refer to footnote [ foot : dlt ] on page for references concerning the notion of a `` dlt pair '' .
we give only a very rough and incomplete idea of the proof of theorem [ thm : relativedifferentialfiltration ] .
to construct the filtration in , one takes the filtration which exists on the open set @xmath414 wherever the morphism @xmath379 is snc , and extends the sheaves to reflexive sheaves that are defined on all of @xmath18 .
it is then not very difficult to show that the sequences ( [ thm : relativedifferentialfiltration].[seq : quotsf ] ) are exact and locally split away from a subset @xmath289 of codimension @xmath415 .
the main point of theorem [ thm : relativedifferentialfiltration ] is , however , that it suffices to remove from @xmath18 a set of codimension @xmath392 . for this , a careful analysis of the codimension - two structure of dlt pairs , cf .
9 ) , proves to be key .
a very important feature of logarithmic differentials is the existence of a residue map . in its simplest form
consider a smooth hypersurface @xmath416 in a manifold @xmath18 .
the residue map is then the cokernel map in the exact sequence @xmath417 more generally , consider a reduced snc pair @xmath331 .
let @xmath418 be any irreducible component and recall from ( * ? ? ?
* ( b ) ) that there exists a residue sequence , @xmath419 & \omega^p_x(\log d ) \ar[r]^(.4){\rho^p } & \omega^{p-1}_{d_0}(\log d_0^c ) \to 0 , } \ ] ] where @xmath420 denotes the `` restricted complement '' of @xmath421 . more generally , if @xmath378 is an snc morphism of @xmath339 we have a relative residue sequence @xmath422 & \omega^p_{x / t}(\log d ) \ar[r]^(.4){\rho^p } & \omega^{p-1}_{d_0/t}(\log d_0^c ) \to 0 . }
\ ] ] the sequence is not a sequence of locally free sheaves on @xmath18 , and its restriction to @xmath421 will never be exact on the left .
however , an elementary argument , cf .
2.13.2 ) , shows that restriction of to @xmath421 induces the following exact sequence @xmath423 which is very useful for inductive purposes .
we recall without proof the following elementary fact about the residue sequence .
[ fact : poletest ]
if @xmath424 is any reflexive form , then @xmath425 if and only if @xmath426 . if the pair @xmath331 is not snc , no residue map exists in general .
however , if @xmath331 is dlt , then ( * ? ? ?
* cor . 5.52 ) applies to show that @xmath421 is normal , and an analogue of the residue map @xmath427 exists for sheaves of reflexive differentials .
( 8,6)(0,0 ) ( 0.0 , 0.0 ) ( -1.65 , 5.05)smooth curve @xmath78 ( 0.8 , 5.0)(2 , -1)1.3 ( 6.5 , 4.7)@xmath18 ( singular surface ) to illustrate the problem we are dealing with , consider a normal space @xmath18 that contains a smooth weil divisor @xmath428 , similar to the one sketched in figure [ fig : sct ] .
one can easily construct examples where the singular set @xmath429 is contained in @xmath78 and has codimension @xmath404 in @xmath18 , but codimension one in @xmath78 . in this setting , a reflexive form @xmath430}_{x}(\log d_0 ) |_{d_0 } \bigr)$ ] is simply the restriction of a logarithmic form defined outside of @xmath292 , and the form @xmath431}(\sigma)$ ] is the extension of the well - defined form @xmath432 over @xmath292 , as a rational form with poles along @xmath433 . if the singularities of @xmath18 are bad , it will generally happen that the extension @xmath431}(\sigma)$ ] has poles of arbitrarily high order . theorem [ thm : relativereflexiveresidue ] asserts that this does not happen when @xmath339 is dlt .
[ thm : relativereflexiveresidue ] let @xmath331 be a dlt pair with @xmath434 and let @xmath435 be an irreducible component .
let @xmath436 be a surjective morphism to a normal variety @xmath393 such that the restricted map @xmath437 is still surjective .
then , there exists a non - empty open subset @xmath394 , such that the following holds if we denote the preimages as @xmath438 , @xmath396 , and the `` complement '' of @xmath439 as @xmath440 . 1 .
[ il : relreflresidue ] there exists a sequence @xmath441}_{x^\circ / t^\circ}(\log ( \lfloor d^\circ\rfloor - d_0^\circ ) ) \to \omega_{x^\circ / t^\circ}^{[r ] } ( \log \lfloor d^\circ \rfloor ) \\
\xrightarrow{\rho^{[r ] } } \omega^{[r-1]}_{d^\circ_0/t^\circ}(\log d_0^{\circ , c } ) \to 0 \end{gathered}\ ] ] which is exact in @xmath135 outside a set of codimension at least @xmath442 .
this sequence coincides with the usual residue sequence wherever the pair @xmath443 is snc and the map @xmath444 is an snc morphism of @xmath443 .
[ il : restrrelreflresidue ] the restriction of sequence ( [ thm : relativereflexiveresidue].[il : relreflresidue ] ) to @xmath421 induces a sequence @xmath441}_{d^\circ_0/t^\circ}(\log d_0^{\circ , c } ) \to
\omega_{x^\circ / t^\circ}^{[r]}(\log \lfloor d^\circ \rfloor ) |_{d_0^\circ } ^ { * * } \\
\xrightarrow{\rho^{[r]}_{d^\circ_0 } } \omega^{[r-1]}_{d^\circ_0/t^\circ}(\log d_0^{\circ , c } ) \to 0 \end{gathered}\ ] ] which is exact on @xmath439 outside a set of codimension at least @xmath404 and coincides with the usual restricted residue sequence wherever the pair @xmath443 is snc and the map @xmath445 is an snc morphism of @xmath446 . as before ,
the proof of theorem [ thm : relativereflexiveresidue ] relies on our knowledge of the codimension - two structure of dlt pairs .
fact [ fact : poletest ] and theorem [ thm : relativereflexiveresidue ] together immediately imply that the residue map for reflexive differentials can be used to check if a reflexive form has logarithmic poles along a given boundary divisor .
[ rem : poletest ] in the setting of theorem [ thm : relativereflexiveresidue ] , if @xmath447}_x(\log \lfloor d \rfloor ) \bigr)$ ] is any reflexive form , then @xmath448}_x(\log \lfloor d \rfloor -d_0 ) \bigr)$ ] if and only if @xmath431}(\sigma ) = 0 $ ] .
let @xmath18 be a smooth variety and @xmath449 a smooth , irreducible divisor .
the first residue sequence of the pair @xmath339 then reads @xmath450 and we obtain a connecting morphism of the long exact cohomology sequence , @xmath451 in this setting , the standard description of the first chern class in terms of the connecting morphism , ( * ? ? ?
7.4 ) , asserts that @xmath452 where @xmath453 is the constant function on @xmath78 with value one .
theorem [ thm : chernclass ] generalises identity to the case where @xmath339 is a reduced dlt pair with irreducible boundary divisor .
[ thm : chernclass ] let @xmath331 be a dlt pair , @xmath454 irreducible .
then , there exists a closed subset @xmath289 with @xmath455 and a number @xmath456 such that @xmath457 is cartier on @xmath458 , such that @xmath459 is smooth , and such that the restricted residue sequence @xmath460}(\log d)|_d^ { * * } \overset{\rho_d}{\longrightarrow } \operatorname{\mathcal o}_{d } \to 0\ ] ] defined in theorem [ thm : relativereflexiveresidue ] is exact on @xmath461 .
moreover , for the connecting homomorphism @xmath462 in the associated long exact cohomology sequence @xmath463 we have @xmath464 the proof of theorem [ thm : generalpullback ] is rather involved .
to illustrate the idea of the proof , we concentrate on a very special case , and give only indications what needs to be done to handle the general setup .
the following simplifying assumptions will be maintained throughout the present section [ ssec : ioppb ] .
[ sass : steenbrink ] the space @xmath18 has dimension @xmath465 .
it is klt , has only one single isolated singularity @xmath466 , and the divisor @xmath78 is empty .
the morphism @xmath353 is a resolution of singularities , whose exceptional set @xmath467 is a divisor with simple normal crossing support . to prove theorem [ thm : generalpullback ]
, we need to show in essence that reflexive differential forms @xmath468}_x \bigr)$ ] pull back to give differential forms @xmath469 .
the following observation , an immediate consequence of fact [ fact : extension1 ] , turns out to be key .
[ obs : xxl ] to give a reflexive differential @xmath468}_x \bigr)$ ] , it is equivalent to give a differential form @xmath470 , defined on the smooth locus of @xmath18 . since the resolution map identifies the open subvarieties @xmath471 and @xmath472 , we see that to give a reflexive differential @xmath468}_x \bigr)$ ] , it is in fact equivalent to give a differential form @xmath473 .
in essence , observation [ obs : xxl ] says that to show theorem [ thm : generalpullback ] , we need to prove that the natural restriction map @xmath474 is in fact surjective . in other words , we need to show that any differential form on @xmath292 , which is defined outside of the @xmath357-exceptional set @xmath475 , automatically extends across @xmath475 , to give a differential form defined on all of @xmath292 .
this is done in two steps .
we first show that the restriction map @xmath476 is surjective .
in other words , we show that any differential form on @xmath292 , defined outside of @xmath475 , extends as a form with logarithmic poles along @xmath475 .
secondly , we show that the natural inclusion map @xmath477 is likewise surjective .
in other words , we show that globally defined differentials forms on @xmath292 , which are allowed to have logarithmic poles along @xmath475 , really do not have any poles .
surjectivity of the morphisms and together will then imply surjectivity of , finishing the proof of theorem [ thm : generalpullback ] . the arguments used to prove surjectivity of and , respectively , are of rather different nature .
we will sketch the arguments in sections [ sec : ext2 ] and [ sec : ext3 ] below . under the simplifying assumptions [ sass : steenbrink ] ,
surjectivity of the map has essentially been shown by steenbrink and van straten , @xcite .
we give a brief synopsis of their line of argumentation and indicate additional steps of argumentation required to handle the general setting . to start , recall from ( * ? ? ?
* iii ex .
2.3e ) that the map is part of the standard sequence that defines cohomology with supports , @xmath478 we aim to show that the last term in vanishes .
there are two main ingredients to the proof : formal duality and steenbrink s vanishing theorem .
[ thm : duality ] under the assumptions [ sass : steenbrink ] , if @xmath170 is any locally free sheaf on @xmath292 and @xmath479 any number , then there exists an isomorphism @xmath480 where @xmath481 denotes completion with respect to the maximal ideal @xmath482 of the point @xmath466 , and where @xmath483 .
a brief introduction to formal duality , together with a readable , self - contained proof of theorem [ thm : duality ] is found in ( * ? ? ?
* appendix a ) while hartshorne s lecture notes @xcite are the standard reference for these matters .
if @xmath86 , @xmath224 are any two numbers with @xmath484 , then @xmath485 .
steenbrink s vanishing theorem is proven using local hodge theory of isolated singularities . for @xmath486 , the sheaves
@xmath487 and @xmath488 are isomorphic . in this case
, the steenbrink vanishing theorem reduces to grauert - riemenschneider vanishing , @xcite . setting @xmath489 and using
that @xmath490 , formal duality and steenbrink vanishing together show that @xmath491 , for @xmath492 , proving surjectivity of for these values of @xmath86 .
the other cases need to be treated separately .
case @xmath493 : : after passing to an index - one cover , surjectivity of in case @xmath494 follows almost directly from the definition of klt , cf .
* sect . 5 ) .
case @xmath495 : : in this case one uses the duality between @xmath496 and the tangent sheaf @xmath497 , and the fact that any section in the tangent sheaf of @xmath18 always lifts to the canonical resolution of singularities , cf .
* sect . 6 ) .
the argument outlined above , using formal duality and steenbrink vanishing , works only because we were assuming that the singularities of @xmath18 are isolated . in the general case , where the simplifying assumptions [ sass : steenbrink ] do not necessarily hold , this is not necessarily the case . in order to deal with non - isolated singularities ,
one applies a somewhat involved cutting - down procedure , as indicated in figure [ fig : tswah ] .
+ @xmath498{ass197 } } \put ( 3.4 , 3.7){\scriptsize divisor $ e_1 $ } \put ( 3.4 , 3.6){\vector(-1 , -1){0.4 } } \put ( 2.9 , 0.3){\scriptsize divisor $ e_0 $ } \put ( 2.8 , 0.4){\vector(-2 , 1){0.4 } } \end{picture } \ar[rr]^{\gamma}_{\text{resolution map } } & & \begin{picture}(4,4)(0,0 ) \put ( 0.0 , 4.2){singular space $ x$ } \put ( 0.0 , 0.2){\includegraphics[height=3.5cm]{family_sing } } \put ( 3.9 , 2.6){\scriptsize point $ \gamma(e_0)$ } \put ( 3.8 , 2.6){\vector(-2 , -1){1.2 } } \put ( 2.4 , 1.87){\scriptsize $ \bullet$ } \put ( 3.8 , 1.3){\scriptsize curve $ \gamma(e_1)$ } \put ( 3.8 , 1.5){\vector(-1 , 1){0.4 } } \end{picture } } \ ] ] the figure sketches a situation where @xmath18 is a threefold whose singular locus is a curve . near the general point of the singular locus
, the variety @xmath18 looks like a family of isolated surfaces singularities .
the exceptional set @xmath475 of the resolution map @xmath357 contains two irreducible divisors @xmath499 and @xmath500 . this way , it is often possible to view non - isolated log canonical singularities a family of isolated singularities , where surjectivity of can be shown on each member of the family .
to conclude that it holds on all of @xmath292 , the following strengthening of steenbrink vanishing is required .
[ thm : omegavanishing ] let @xmath331 be a log canonical pair of dimension @xmath501 .
if @xmath502 is a log resolution of singularities with exceptional set @xmath475 and @xmath503 then @xmath504 for all @xmath505 .
the proof of theorem [ thm : omegavanishing ] essentially relies on the fact that log canonical pairs are du bois , @xcite .
the du bois property generalises the notion of rational singularities . for an overview , see @xcite .
let @xmath506 be any differential form on @xmath292 that is allowed to have logarithmic poles along @xmath475 .
to show surjectivity of the inclusion map , we need to show that @xmath507 really does not have any poles along @xmath475 .
to give an idea of the methods used to prove this , we consider only the case where @xmath508 .
we discuss two particularly simple cases first .
assume that @xmath475 is irreducible . to show that @xmath507 does not have any logarithmic poles along @xmath475 , recall from fact [ fact : poletest ] that it suffices to show that @xmath507 is in the kernel of the residue map @xmath509 on the other hand , we know from a result of hacon - mckernan , ( * ? ? ?
1.5(2 ) ) , that @xmath475 is rationally connected , so that @xmath510 .
this clearly shows that @xmath507 is in the kernel of @xmath427 and completes the proof when @xmath475 is irreducible . in general , the divisor @xmath475 need not be irreducible .
let us therefore consider the next difficult case where @xmath475 is reducible with two components , say @xmath512 .
the resolution map @xmath357 will then factor via a @xmath357-relative minimal model program of the pair @xmath513 , which we assume for simplicity to have the following particularly special form , sketched is partially taken from @xcite . ] also in figure [ fig : srklt ] .
+ @xmath514{d2-secondblowup } } \put ( 0.0 , 4.2){snc surface pair $ ( z , e_1+e_2)$ } \put ( 0.2 , 3.4){\scriptsize divisor $ e_1 $ } \put ( 1.5 , 3.4){\vector(4 , -1){0.6 } } \put(-0.5 , 1.2){\scriptsize divisor $ e_2 $ } \put ( 0.8 , 1.2){\vector(4 , -1){1.1 } } \end{picture } \quad \ar[rr]^(.54){\lambda_1}_(.54){\text{contracts $ e_1 $ } } \ar@<-10mm>@/_5mm/[drr]_(.55){\text{resolution map\ } } ^(.55){\gamma } & & \begin{picture}(4,4)(0,0 ) \put ( 0.0 , 0.2){\includegraphics[height=3.5cm]{d2-firstblowup } } \put ( 0.0 , 4.2){dlt surface pair $ ( z_1 , e_{2,1})$ } \put ( 2.8 , 1.4){\scriptsize divisor $ e_{2,1}$ } \put ( 2.7 , 1.6){\vector(-1 , 4){0.2 } } \end{picture } \ar[d]_(.55){\lambda_2}^(.55){\text{contracts $ e_{2,1}$ } } \\ & & \begin{picture}(3.5 , 3.2)(0,0 ) \put ( 0.0 , 0.0){\includegraphics[width=3.5cm]{d2 } } \put ( 0.0 , 2.5){klt surface $ x$ } \end{picture } } \ ] ] this sketch shows a resolution of an isolated klt surface singularity , and the decomposition of the resolution map given by the minimal model program of the snc pair @xmath515 . @xmath516^(.55){\lambda_1}_(.55){\text{contracts $ e_1 $ to a point } } & & & & z_1 \ar[rrrrr]^{\lambda_2}_{\text{contracts $ e_{2,1 } : = ( \lambda_1)_*(e_2)$ to a point } } & & & & & x. } \ ] ] in this setting , the arguments outlined above apply verbatim to show that @xmath507 has no poles along the divisor @xmath500 . to show that @xmath507 does not have
any poles along the remaining component @xmath517 , observe that it suffices to consider the induced reflexive form on the possibly singular space @xmath518 , say @xmath519}_{z_1 } ( \log e_{2,1 } ) \bigr)$ ] , where @xmath520 , and to show that @xmath521 does not have any poles along @xmath522 .
for that , we follow the same line of argument once more , accounting for the singularities of the pair @xmath523 .
the pair @xmath524 is dlt , and it follows from adjunction that the divisor @xmath522 is necessarily normal , ( * ? ? ?
* cor . 5.52 ) .
using the residue map for reflexive differentials on dlt pairs that was constructed in theorem [ thm : relativereflexiveresidue ] , @xmath525 } : h^0\bigl ( x_1,\ , \omega^{[p]}_{z_1}(\log e_{2,1 } ) \bigr ) \to h^0\bigl ( e_{2,1},\ , \omega^{[p-1]}_{e_{2,1 } } \bigr),\ ] ] we have seen in remark [ rem : poletest ] that it suffices to show that @xmath431}(\sigma_1 ) = 0 $ ] . because the morphism @xmath526 contracts the divisor @xmath522 to a point , the result of hacon - mckernan will again apply to show that @xmath522 is rationally connected . even though there are numerous examples of rationally connected spaces that carry non - trivial reflexive forms , we claim that in our special setup we do have the vanishing @xmath527}_{e_{2,1 } } \bigr ) = 0.\ ] ] for this , recall from the adjunction theory for weil divisors on normal spaces , ( * ? ? ?
* chapt . 16 and prop . 16.5 ) and ( * ? ? ?
3.9 and glossary ) , that there exists a weil divisor @xmath528 on the normal variety @xmath522 which makes the pair @xmath529 klt .
now , if we knew that the extension theorem would hold for the pair @xmath529 , we can prove the vanishing statement , arguing exactly as in the proof of corollary [ cor : kltrcc ] , where we show the non - existence of reflexive forms on rationally connected klt spaces as a corollary of the pull - back theorem [ thm : generalpullback ] .
since @xmath530 , this suggests an inductive proof , beginning with easy - to - prove extension theorems for reflexive forms on surfaces , and working our way up to higher - dimensional varieties .
the proof in @xcite follows this inductive pattern . to handle the general case , where the simplifying assumptions [ sass : steenbrink ] do not necessarily hold true
, we need to work with pairs @xmath339 where @xmath78 is not necessarily empty , the @xmath357-relative minimal model program might involve flips , and the singularities of @xmath18 need not be isolated .
all this leads to a slightly protracted inductive argument , heavily relying on cutting - down methods and outlined in detail in ( * ? ? ?
19 ) . in view of the viehweg - zuo construction
, it would be very interesting to know if a variant of the pull - back theorem [ thm : generalpullback ] holds for symmetric powers of @xmath281}_x(\log d)$ ] , or for other tensor powers . as shown by examples , cf .
3.1.3 ) , the nave generalisation of theorem [ thm : generalpullback ] is wrong .
still , it seems conceivable that a suitable generalisation , perhaps formulated in terms of campana s orbifold differentials , might hold .
however , note that several of the key ingredients used in the proof of theorem [ thm : generalpullback ] , including steenbrink s vanishing theorem , rely on ( local ) hodge theory , for which no version is known for tensor powers of differential forms .
is there a formulation of the pull - back theorem [ thm : generalpullback ] that holds for symmetric and other tensor powers of differential forms ?
examples suggest that the pull - back theorem [ thm : generalpullback ] is optimal , and that the class of log canonical pairs is the natural class of spaces where a pull - back theorem can hold .
to what extend is the pull - back theorem [ thm : generalpullback ] optimal ?
is there a version of the pull - back theorem that does not require the log canonical divisor @xmath531 to be @xmath77-cartier ? if we are interested only in special values of @xmath86 , is the divisor @xmath19 the smallest possible
? the last question concerns the generalisation of the bogomolov - sommese vanishing theorem .
one of the main difficulties with its current formulation is the requirement that the sheaf @xmath244 be @xmath77-cartier .
we have seen in section [ ssec : motiv ] how interesting reflexive subsheaves @xmath532}_x$ ] can often be constructed using the harder - narasimhan filtration . unless the space @xmath18 is @xmath77-factorial , there is , however , no way to guarantee that a sheaf constructed this way will actually be @xmath77-cartier . the property to be @xmath77-factorial , however , is not stable under taking hyperplane sections and difficult to guarantee in practise .
is there a version of the generalised bogomolov - sommese vanishing theorem , corollary [ cor : bs ] , that does not require the sheaf @xmath244 to be @xmath77-cartier ?
we conclude this paper by sketching a proof of the viehweg conjecture [ conj : viehweg ] in one special case , illustrating the use of the methods introduced in sections [ sec : vz ] and [ sec : ext ] . as in section [ ssec : motiv ] we consider a family @xmath8 of canonically polarised varieties over a quasi - projective threefold . assuming that @xmath7 is of maximal variation , we would like to show that the logarithmic kodaira dimension @xmath29 can not be zero . [
prop : vzsimple2 ] let @xmath8 be a smooth , projective family of canonically polarised varieties over a smooth , quasi - projective base manifold of dimension @xmath533 .
assume that the family @xmath7 is of maximal variation , i.e. , that @xmath43 . then @xmath257 .
the proof of proposition [ prop : vzsimple2 ] follows the line of argumentation outlined in section [ ssec : motiv ] .
we prove that the picard number of a suitable minimal model can not be one , thereby exhibiting a fibre space structure to which induction can be applied .
the presentation follows ( * ? ? ?
* sect . 9 ) .
in essence , we follow the line of argument sketched in section [ ssec : motiv ] .
we argue by contradiction , i.e. , we maintain the assumptions of proposition [ prop : vzsimple2 ] and assume in addition that @xmath258 . as before
, choose a smooth compactification @xmath253 such that @xmath534 is a divisor with only simple normal crossings .
let @xmath52 be the rational map obtain by a run of the minimal model program for the pair @xmath48 and set @xmath535 .
the following is then known to hold . 1 .
the variety @xmath276 is normal and @xmath77-factorial .
the variety @xmath276 is log terminal .
the pair @xmath536 is dlt .
3 . there exists a number @xmath537 such that @xmath538 .
in particular , the divisor @xmath539 is numerically trivial .
by viehweg - zuo s theorem [ thm : vz ] , there exists a number @xmath102 and a big invertible sheaf @xmath540 . as we have seen in proposition [
lem : pushdowna ] , this induces a reflexive sheaf @xmath541 } \omega^1_{y_\lambda } ( \log d_\lambda)$ ] of rank one and kodaira - iitaka dimension @xmath542 .
[ [ the - harder - narasimhan - filtration - of - omega1_y_lambdalog - d_lambda ] ] the harder - narasimhan filtration of @xmath281}_{y_\lambda}(\log d_\lambda)$ ] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ for simplicity , we prove claim [ claim:5 - 4 ] only in case where the canonical divisor @xmath544 is cartier , and where the space @xmath276 therefore has only canonical singularities . for a proof in the general setup , the same line of argumentation applies after passing to a global index - one cover .
we argue by contradiction and assume that @xmath545 . as before ,
let @xmath546 be a general complete intersection curve in the sense of mehta - ramanathan , cf .
since the general complete intersection curve @xmath10 avoids the singular locus of @xmath276 , we obtain that the restricted sheaf of khler differentials @xmath547 as well as its dual @xmath548 , the restriction of the tangent sheaf , are locally free .
further , the numerical triviality of @xmath549 implies that @xmath550}_{y_\lambda}(\log d_\lambda ) \bigr).c = c_1\bigl(\operatorname{sym}^{[m ] } \omega^1_{y_\lambda}(\log d_\lambda ) \bigr).c = 0.\ ] ] on the other hand , since @xmath551 is big , we have that @xmath552 . as in the proof of proposition
[ prop : vzec1 ] , this implies that the restricted sheaves @xmath547 as well as its dual @xmath548 , are not semistable .
the maximal destabilising subsheaf of @xmath548 is semistable and of positive degree , hence ample . in this setup , a variant ( * ? ? ?
* cor . 5 ) of miyaoka s uniruledness criterion ( * ? ? ?
8.6 ) applies to give the uniruledness of @xmath276 . for more details on this criterion , see the survey @xcite .
to finish the argument , let @xmath553 be a resolution of singularities .
since uniruledness is a birational property , the space @xmath554 is uniruled and therefore has kodaira - dimension @xmath555 . on the other hand ,
since @xmath276 has only canonical singularities , the @xmath359-linear equivalence class of the canonical bundle @xmath556 is given as @xmath557 but because @xmath544 is @xmath77-linearly equivalent to the trivial divisor , we obtain that @xmath558 , a contradiction . claim [ claim:5 - 4 ] implies that @xmath559 and it follows that for any rational number @xmath560 , @xmath561 now choose one @xmath562 and run the log minimal model program for the dlt pair @xmath563 .
this way one obtains morphisms and birational maps as follows @xmath564^{\mu}_{\text{minimal model program } } & & & y_\mu \ar[rrr]^{\pi}_{\text{mori fibre space } } & & & z. } \ ] ] again , let @xmath565 be the cycle - theoretic image of @xmath543 .
the main properties of @xmath566 and @xmath567 are summarised as follows . 1 .
the variety @xmath566 is normal and @xmath77-factorial .
the variety @xmath566 is log terminal .
the pair @xmath568 is dlt .
the divisor @xmath569 is numerically trivial .
there exists a reflexive sheaf @xmath570 } \omega^1_{y_\mu } ( \log d_\mu)$ ] of rank one and kodaira - iitaka dimension @xmath571 .
since @xmath573 , some positive multiples of @xmath544 and @xmath574 are numerically equivalent . for any two rational numbers @xmath575 , the divisors @xmath576 and @xmath577 are thus again numerically equivalent up to a positive rational multiple .
the birational map @xmath114 is therefore a minimal model program for the pair @xmath578 , independently of the number @xmath562 chosen in its construction .
it follows that @xmath579 is a limit of dlt pairs and therefore log canonical . as before
, let @xmath583 be a general complete intersection curve .
again , the existence of the viehweg - zuo sheaf @xmath584 implies that the sheaf of reflexive differentials @xmath281}_{y_\mu}(\log d_\mu)$ ] is not semistable , and contains a destabilising subsheaf @xmath585}_{y_\mu}(\log d_\mu)$ ] with @xmath586 . since the intersection number @xmath587 is positive , the rank @xmath270 of the sheaf @xmath588 must be strictly less than @xmath589 , and its determinant is a subsheaf of the sheaf of logarithmic @xmath270-forms , @xmath590}_{y_\mu}(\log d_\mu ) \quad\text{with}\quad c_1(\det \mathcal b_\mu).c > 0 \quad\text{and}\quad r<\dim y_\mu.\ ] ] if @xmath591 , then the sheaf @xmath592 would necessarily be @xmath77-ample , violating the bogomolov - sommese vanishing theorem for log canonical pairs , corollary [ cor : bs ] .
this finishes the proof of claim [ claim:55 ] .
now , if @xmath593 is a general fibre of @xmath367 and @xmath594 , then @xmath22 is a normal curve or surface , and the pair @xmath595 is log canonical and has kodaira dimension @xmath596 . by (
4.11 ) , the variety @xmath22 is even @xmath77-factorial .
it is then possible to argue by induction : assuming that viehweg s conjecture holds for families over surfaces , one obtains that the restriction of the family @xmath7 to the strict transform @xmath597 can not be of maximal variation .
since the fibres dominate the variety , this contradicts the assumption that the family @xmath7 is of maximal variation , and therefore finishes the sketch of proof of proposition [ prop : vzsimple2 ] . : _ introduction to hodge theory _ , smf / ams texts and monographs , vol .
8 , american mathematical society , providence , ri , 2002 , translated from the 1996 french original by james lewis and peters . 1924513 ( 2003g:14009 ) : _ effective bounds for semipositive sheaves and for the height of points on curves over complex function fields _ , compositio math .
* 76 * ( 1990 ) , no . 1 - 2 , 6985 , algebraic geometry ( berlin , 1988 ) .
mr1078858 ( 91m:14038 ) : _ lectures on vanishing theorems _ , dmv seminar , vol .
20 , birkhuser verlag , basel , 1992 .
available from the author s web site at http://www.uni-due.de/~mat903/books.html[http://www.uni-due.de/@xmath598mat903/books.html ] .
mr1193913 ( 94a:14017 ) : _ extension theorems for differential forms , and bogomolov - sommese vanishing on log canonical varieties _ ,
compos . math .
* 146 * ( 2010 ) , 193219 , http://dx.doi.org/10.1112/s0010437x09004321[doi:10.1112/s0010437x09004321 ] . : _ differential forms on log canonical spaces _ ,
publ . math .
ihes . 114 ( 2011 ) , no .
1 , 87169 .
http://dx.doi.org/10.1007/s10240-011-0036-0[doi:10.1007/s10240-011-0036-0 ] .
an extended version with additional graphics is available as http://arxiv.org/abs/1003.2913[arxiv:1003.2913 ] . : _ algebraic geometry _ , graduate texts in mathematics , vol .
76 , springer - verlag , new york , 1982 , an introduction to birational geometry of algebraic varieties , north - holland mathematical library , 24 .
mr637060 ( 84j:14001 ) : _ families of varieties of general type over compact bases _ , adv . math .
* 218 * ( 2008 ) , no . 3 , 649652 , http://dx.doi.org/10.1016/j.aim.2008.01.005[doi:10.1016/j.aim.2008.01.005 ] .
mr2414316 ( 2009d:14042 ) : _ rational curves on algebraic varieties _ , ergebnisse der mathematik und ihrer grenzgebiete .
3 . folge . a series of modern surveys in mathematics , vol . 32 , springer - verlag , berlin , 1996 .
mr1440180 ( 98c:14001 ) : _ birational geometry of algebraic varieties _ , cambridge tracts in mathematics , vol .
134 , cambridge university press , cambridge , 1998 , with the collaboration of c. h. clemens and a. corti , translated from the 1998 japanese original .
2000b:14018 : _ flips and abundance for algebraic threefolds _ , socit mathmatique de france , paris , 1992 , papers from the second summer seminar on algebraic geometry held at the university of utah , salt lake city , utah , august 1991 , astrisque no .
211 ( 1992 ) .
mr1225842 ( 94f:14013 ) : _ subvarieties of moduli stacks of canonically polarized varieties _ ,
proceedings of the ams summer research institute held at the university of washington , seattle , wa , july 25august 12 , 2005 , proceedings of symposia in pure mathematics , american mathematical society , providence , ri , 2006 , to appear . : _ young person s guide to moduli of higher dimensional varieties _ , algebraic geometry
seattle 2005 .
part 2 , proc .
pure math .
80 , amer .
soc . , providence , ri , 2009 , pp .
mr2483953 ( 2009m:14051 ) : _ positivity in algebraic geometry .
, ergebnisse der mathematik und ihrer grenzgebiete .
3 . folge . a series of modern surveys in mathematics [ results in mathematics and related areas .
3rd series .
a series of modern surveys in mathematics ] , vol .
49 , springer - verlag , berlin , 2004 , positivity for vector bundles , and multiplier ideals .
mr2095472 : _ algebraic geometry and arithmetic curves _ , oxford graduate texts in mathematics , vol .
6 , oxford university press , oxford , 2002 , translated from the french by reinie ern , oxford science publications . mr1917232 ( 2003g:14001 ) : _ deformations of a morphism along a foliation and applications _ ,
algebraic geometry , bowdoin , 1985 ( brunswick , maine , 1985 ) , proc . sympos .
pure math .
46 , amer .
soc . , providence , ri , 1987 , pp .
mr927960 ( 89e:14011 ) : _ young person s guide to canonical singularities _ , algebraic geometry , bowdoin , 1985 ( brunswick , maine , 1985 ) , proc .
pure math .
46 , amer .
providence , ri , 1987 , pp .
mr927963 ( 89b:14016 ) : _ algebraic number fields _ , proc .
mathematicians ( stockholm , 1962 ) , inst .
mittag - leffler , djursholm , 1963 , english translation : amer .
( 2 ) * 31 * ( 1963 ) , 2539 , pp .
mr0202709 ( 34 # 2569 ) : _ weak positivity and the additivity of the kodaira dimension for certain fibre spaces _ , algebraic varieties and analytic varieties ( tokyo , 1981 ) , adv . stud .
pure math .
, vol . 1 , north - holland , amsterdam , 1983 , pp . 329353 .
mr715656 ( 85b:14041 ) : _ quasi - projective moduli for polarized manifolds _ , ergebnisse der mathematik und ihrer grenzgebiete ( 3 ) [ results in mathematics and related areas ( 3 ) ] , vol .
30 , springer - verlag , berlin , 1995 .
mr1368632 ( 97j:14001 ) : _ positivity of direct image sheaves and applications to families of higher dimensional manifolds _ , school on vanishing theorems and effective results in algebraic geometry ( trieste , 2000 ) , ictp lect .
notes , vol . 6 , abdus salam int . cent
. theoret .
phys . , trieste , 2001 , available on the http://www.ictp.trieste.it/~pub_off/services[ictp web site ] , pp .
mr1919460 ( 2003f:14024 ) : _ hodge theory and complex algebraic geometry .
i _ , english ed . , cambridge studies in advanced mathematics , vol . 76 , cambridge university press , cambridge , 2007 , translated from the french by leila schneps .
mr2451566 ( 2009j:32014 ) | the shafarevich hyperbolicity conjecture , proven by arakelov and parshin , considers a smooth , projective family of algebraic curves over a smooth quasi - projective base curve @xmath0 .
it asserts that if @xmath0 is of special type , then the family is necessarily isotrivial .
this survey discusses hyperbolicity properties of moduli stacks and generalisations of the shafarevich hyperbolicity conjecture to higher dimensions .
it concentrates on methods and results that relate moduli theory with recent progress in higher dimensional birational geometry . # |
variable stars are useful tools to improve our understanding on stellar evolution and structure of stars .
star clusters are unique laboratories to study the stellar evolution as star clusters provide a sample of stars having same age , distance , initial composition and spanning range of masses .
additionally , the light curves of several stars can be produced simultaneously to know the real variable stars among them .
studies of variable stars in star clusters are carried out since several years ( herbst et al .
1994 ; edwards et al . 1993 ; oliveira & casali 2008 ; grankin et al .
star clusters have different types of variable stars as stars show variability at various stages of their evolutionary phases . in the present study
, we made a search for variable stars in an extremely young open cluster berkeley 59 ( be 59 ) .
be 59 is a young cluster and it contains a significant numbers of pre - main sequence ( pms ) stars i.e , stars evolving from their birthline to the main - sequence ( ms ) .
this cluster ( @xmath6 = 00@xmath7 02@xmath8 13@xmath9 , @xmath10 = + 67@xmath11 25@xmath12 11@xmath13 , l = [email protected] , b = [email protected] ) is located at the centre of ob4 stellar association surrounded by hii region , sharpless region s171 ( yang & fukui 1992 ) .
previous studies have revealed that be 59 consists of many pms stars having age around 2 myr ( pandey et al .
be 59 was first studied photographically by blanco & williams ( 1959 ) .
they concluded that stars of spectral types o7 to b5 are still surrounded by a part of the parent molecular cloud .
a multiwavelength study on be 59 was presented by pandey et al .
( 2008 , hereafter p08 ) .
they found that this cluster is located at a distance of 1.0 kpc and having a variable extinction of @xmath14=1.4 - 1.8 mag .
they also found that around 32% of the h@xmath15 emission stars in be 59 exhibit nir excess indicating that they still have their inner disk . during a campaign of photometric monitoring , majaess et al .
( 2008 ) presented the nature of the three eclipsing systems ( bd+66@xmath111673 , 2mass 00104558 + 6127556 and 2msass 19064659 + 4401458 ) in the field of be 59 using @xmath16 photometry and spectroscopic observations . from the above studies it is clear that be 59 contains high , intermediate and low mass pms stars .
pms objects are , in general , classified into t tauri stars ( ttss ) and herbig ae / be stars
. the ttss are newly formed low - mass stars ( @xmath173m@xmath18 ) which are contracting towards the ms ( herbig 1977 ) , while stars of intermediate mass ( @xmath193 - 10 m@xmath20 ) are known as herbig ae / be stars ( herbig 1960 ; finkenzeller & mundt 1984 ; strom et al .
ae / be stars are either contracting towards ms or they have reached the ms .
the ttss can be classified further as weak line ttss ( wttss ) and classical ttss ( cttss ) on the basis of the strength of the h@xmath15 emission lines ( strom et al .
the strength of the emission line is measured by its equivalent width ( ew ) . the wttss exhibit a weak , narrow h@xmath21 ( ew@xmath22 ) in emission with no or little infrared excess , while cttss generally display strong h@xmath15 emission line ( ew@xmath2310@xmath24 ) , large ultraviolet and infrared excess . the large h@xmath21 emission ( ew@xmath2310@xmath24 )
is believed to be a result of the increased mass accretion onto the star ( e.g. cabrit et al .
1990 ; calvet & hartmann 1992 ) .
it is found that pms stars show strong variability both in photometric and spectroscopic ( van den ancker et al . 1998 ; catala et al .
1999 ) observations .
this could be due to the presence of circumstellar obscuration and spots present on the photosphere .
in fact variability is a defining characteristics of ttss ( herbst et al .
1994 ; scholz et al .
2009 and references therein ) .
both wttss and cttss show variation in their brightness .
these variations are found to occur at all wavelengths , from x - ray to infrared .
variability time scale of ttss ranges from few minutes to years ( appenzeller & mundt 1989 ) . the photometric variations are believed to originate from several mechanisms like circumstellar gas and dust ( remnant of parent molecular cloud ) , accretion and magnetic field ( herbst et al . 1994 ) .
the variations in the brightness of ttss are most probably due to the presence of spots ( cool or hot ) at stellar surface and circumstellar disk .
the cool spots on the surface of the stars are produced by the emergence of stellar magnetic fields on the photosphere , and is thus an indicator of magnetic activity .
as these spots are present on the photosphere , they rotate with the star .
if these spots are symmetrically distributed over the photosphere , periodic variations are seen in the light curves of the stars .
spot configurations remain constant over time - scales of several days , but undergo changes on time - scales of weeks ( hussain 2002 ) .
the cool spots on the photosphere of star are responsible for brightness variation in wttss , and these objects are found to be fast rotators as they have either thin or no circumstellar disk . therefore the variability in these objects may be mainly due to the brightness variations on the photosphere modulated by the stellar rotation ( messina et al .
2010 ) .
the hot spots on the surface of young stars are supposed to be formed on the surface by infalling gas and are thus a direct consequence of accretion ( lynden - bell & pringle 1974 ; koenigl 1991 ; shu et al .
irregular or non periodic variations are produced because of changes in the accretion rate .
the time scales of varying brightness range from hours to years .
the hot spots cover a smaller fraction of the stellar surface but higher temperature causes larger brightness variations ( carpenter et al .
thus , accreting cttss which are surrounded by circumstellar accretion disk and possess hot spots produced by accreting material from disk on to the star show complex behaviour in their optical and near infrared ( nir ) light curves ( scholz et al .
2009 ) .
ae / be stars also show variability as they move across the instability region in the hertzsprung - russell ( hr ) diagram on their way to the ms .
the right combination of mass , temperature and luminosity of these stars make them to pulsate .
several pms pulsators have been detected until now ( breger 1972 ; kurtz & marang 1995 ; marconi et al . 2000 ; donati et al .
1997 ; zwintz et al .
2005 , 2009 ; zwintz & weiss 2006 ) .
catala ( 2003 ) presented detailed review on the pms pulsations and expected that pms stars could pulsate as @xmath25 scuti stars , however it is necessary to confirm their pms nature because these stars are found above the ms where it is very difficult to say whether they are pms or post ms stars .
but in cluster environment where the age of the cluster is constrained well , the stars found towards the cluster region could be pms stars .
a search of photometric variables in young open cluster be 59 , which is found to have a rich population of pms stars with a large mass range could give us a better insight of the variability time - scales , the amplitude of brightness variation etc . with this aim
, we observed the cluster be 59 at different epochs from 2006 to 2010 .
the paper is organised in the following manner . in section 2 , we describe the observations and data reduction techniques .
we discuss membership of variable stars on the basis of photometric observations in section 3 . in section 4 ,
we determine periods of variable stars .
we discuss variability characteristics as well as period and amplitude distribution of variable stars in section 5 , while in section 6 we conclude our studies .
.log of optical photometric ccd observations . ` exp . ' and ` n ' refer to exposure time and number of frames respectively . [
tab : obslog ] [ cols="<,<,^,^ " , ] @xmath26:foreground ms + in order to find out the possible causes of photometric variations in the identified variable stars in the region of be 59 , it is important to ascertain their membership . in the absence of spectroscopic observations , we established photometric membership on the basis of @xmath27 colour - magnitude and ( @xmath28 colour - colour diagram in next section .
the colour - magnitude diagram ( cmd ) of the cluster region clearly reveals contamination due to field star population ( cf .
figure 7 by p08 ) .
p08 found that the cluster is located at a distance of [email protected] kpc and has a differential reddening @xmath14=1.4 - 1.8 mag .
they also identified a foreground field star population at a distance of @xmath3470 pc with a @xmath14=0.40 mag ( cf .
figure 4 of p08 ) .
therefore , the sample of of identified variables may be contaminated by foreground population .
we have used @xmath30 , @xmath31 cmd and nir @xmath32 colour - colour diagram to find out the association of the identified variables with the cluster be 59 .
5 displays @xmath30 , @xmath31 cmd for the stars observed in the cluster .
the identified variables are marked by triangles ( cttss ) and filled circles ( wttss ) ( cf .
section 3.2 ) . the zero - age main - sequence ( zams ) by girardi et al .
( 2002 ) corrected for the distance of 1.0 kpc and @xmath14 = 1.45 mag is shown by thick continuous line .
the pre - main - sequence ( pms ) isochrones for 0.1 , 1 , 5 and 10 myrs by siess et al .
( 2000 ) have also been overplotted .
the zams corrected for a distance of 470 pc and @xmath14=0.40 mag for the foreground population is shown by thick dashed curve .
5 . reveals that the stars v1 , v5 , v14 , v18 , v21 , v25 , v29 , v30 , v35 , v41 and v42 ( shown with open circles ) follow the zams for field stars , hence could be ms foreground population .
it is worthwhile to point that majority of the field stars ( 10 out of 11 ) are located outside the core ( 1.4 arcmin , p08 ) of the cluster . in order to determine the ages and masses of the probable pms variable stars we have used pms isochrones and evolutionary tracks by siess et al .
( 2000 ) , respectively .
majority of the probable pms cluster members have masses in the range [email protected] and ages in the range 1@xmath17age@xmath175 myrs .
this indicates that the young stellar objects associated with the cluster could be ttss .
star v23 is the brightest variable in the sample and lies near 0.1 myr pms isochrone .
the nir colour - colour diagram ( ccd ) can be used to distinguish the cttss and wttss .
6 shows nir ccd for the identified variables .
the nir data have been taken from the 2mass point source catalogues ( psc ; cutri et al .
the 2mass counterpart of the variables were searched using a match value of 3 arcsec .
of the 42 detected variables , we found 2mass counterparts for 41 stars ( 30 cluster members and 11 foreground ms stars ) .
the 2mass counterparts for v32 star could not be found .
it is the reddest variable star in the sample and lying above 0.1 myr pms isochrone in @xmath34 cmd .
however , on the basis of the location of this star on @xmath34 cmd , we have considered v32 as a pms object . the 2mass magnitudes and colours
were transformed to the cit system using the relations given on the 2mass website .
the solid and long dashed lines represent unreddened ms and giant branch ( bessell & brett 1988 ) , respectively .
the dotted line indicates intrinsic locus of cttss ( meyer et al .
the parallel dashed lines are the reddening vectors drawn from the tip ( spectral type m4 ) of the giant branch ( ` upper reddening line ' ) , from the base ( spectral type a0 ) of the main - sequence branch ( ` middle reddening line ' ) and from the tip of the intrinsic ctts line ( ` lower reddening line ' ) .
the extinction ratios @xmath35 and @xmath36 have been adopted from cohen et al .
we classified the sources into three regions in the nir ccd ( cf .
ojha et al .
the ` f ' sources are those that are located between the upper and middle reddening lines and are considered to be either field stars ( main - sequence stars , giants ) or class iii and class ii sources with small nir excesses .
the identification of class ii sources having a small amount of nir excess is difficult on the basis of ( @xmath28 ccd .
however , recent studies , where class ii sources were identified on the basis of mid - ir observations through spitzer telescope , manifest that these class ii objects lie in ` f ' region near the middle reddening vector ( e.g. chauhan et al . 2011 ) .
the ` t ' sources are those that are located between the middle and lower reddening lines .
these sources are considered to be mostly cttss ( class ii objects ) . `
p ' sources are those that are located in the region redward of the ` t ' region and are most likely class i objects ( protostar - like objects ; ojha et al .
2004b ) .
majority of the stars lying in the ` f ' region are considered as class iii stars / wttss and are shown by filled circles
. however , there seems segregation of sources between @xmath37 @xmath19 1.0 mag and @xmath38 @xmath19 0.50 mag in @xmath2 colour - colour diagram . as discussed above that class ii sources with small amount of nir excess could be found in ` f ' region near the middle reddening vector .
the stars having @xmath37 @xmath19 1.0 mag and @xmath38 @xmath19 0.50 mag are considered ctts and these are shown by filled triangles .
the foreground ms stars as identified from @xmath39 cmd are distributed around the zams and shown by open circles .
the nature of the identified variables is mentioned in the table 2 .
the membership of the variable stars is determined on the basis of their location in @xmath1 cmd and nir ccd .
however , spectroscopic follow up of these objects is needed to confirm their nature .
we used the lomb - scargle ( ls ) periodogram ( lomb 1976 ; scargle 1982 ) to determine the most likely period of a variable star .
the ls method is useful for finding significant periodicities even with unevenly sampled data .
we used the algorithm publicly available at the starlink software database , and verified the periods further with the software period04 ( lenz & breger 2005 ) .
the software period04 provides the frequency and semi - amplitude of the variability in a light curve .
periods derived from the ls method and period04 generally matched well . in order to further verify the period estimates we also used periodogram analysis available online .
for any star showing a spurious period , we visually inspected the phased light curve for that particular period .
the phased light curves for cttss , wttss and field stars are shown in figs . 7 , 8 and 9 , and 10 respectively . figs . 7 , 8 , 9 and 10 plot the mean of differential magnitudes in 0.04 phase bin as a function of phase .
the error bars represent the @xmath40 of the mean . in subsequent section
we discuss the variability characteristics of individual variable star .
light curve of one wtts variable named v38 displayed in fig .
9 does not show periodic variability . this star could be either irregular variable or the period estimated in the present work could be in error . the short term periodic variations ( 2 @xmath3 10 days ) in ttss
are believed to occur due to the axial rotation of a star with an inhomogeneous surface , having either hot or cool spots ( herbst et al .
1987 , 1994 ) .
herbst et al . ( 1994 ) identified three types of tt variables based on their variability timescales ranging from a day to weeks .
type i variability , most often seen in wttss , is characterized by smaller stellar flux variations ( a few times of 0.1 mag ) and results from the rotation of a cool spotted photosphere .
type ii variations have larger brightness variations ( up to @xmath32 mag ) , most often irregular but sometime periodic , are associated with short - lived accretion related hot spots at the stellar surface of cttss .
the rare type iii variations are characterized by luminosity dips lasting from a few days upto several months , which presumably result from circumstellar dust obscuration .
the sample of ttss in the cluster be 59 indicates that @xmath390% ttss have periods less than 15 days .
the estimated periods of 8 cttss vary in the range of 0.652 day to 10.317 day , however stars v3 and v10 have significantly larger periods ( 89.8 day and 58.7 day respectively ) .
the period estimates for wttss vary from 0.109 day to 5.671 day , however @xmath395% wttss have periods @xmath41 2 day . the shorter periods in the case of wttss could be explained as most of the stars at this age might have unlocked from their disks while contracting towards ms .
they spin up because of decreasing radii in order to conserve their angular momentum ( strom et al .
1989 ; skrutskie et al .
1990 ; haisch et al .
2001 ; bouvier et al . 1994 ; herbst & mundt 2005 ; lamm et al .
2005 ; kundurthy et al .
but nordhagen et al .
( 2006 ) have different view and suggest that environment also plays a crucial role in establishing the rotation periods besides age and mass .
the presence of ` o ' type stars in the vicinity would speed up the process of the dispersal of the circumstellar disks causing the stars to spin up at a much younger age .
the amplitude of wttss ranges from 0.022 to 0.273 mag .
the amplitude in the case of cttss has a range of 0.215 to 1.122 mag .
the present results regarding amplitudes are in agreement with grankin et al .
( 2008 for wttss ) and grankin et al .
( 2007 for cttss ) .
the brightness of cttss is found to vary with larger amplitude in comparison to the wttss .
the large amplitude in the case of cttss indicates presence of the hot spots on their surfaces .
the smaller amplitude in the case of wttss suggests that these pms stars might be in the process of dissipation of their circumstellar disk .
the star v3 , a probable ctts , seems to be an interesting object .
it has relatively high excess ( @xmath42 ) and shows relatively high extinction ( cf .
it has a period of 89.80 days with amplitude variation of 1.0 mag and its mass is estimated to be @xmath30.9 @xmath4 .
the relatively high extinction of this object could be due to the presence of the circumstellar disk .
edwards et al .
( 1993 ) found that the stars which showed the evidence of circumstellar disks , inferred from their @xmath38 colours , are slow rotators in comparison to those that lacked circumstellar disks .
relatively high extinction and longer period of v3 support this argument .
this star may belong to type iii as classified by herbst et al .
( 1994 ) , where relatively larger luminosity dips for longer period are expected due to circumstellar dust obscuration .
the h@xmath21 emission star v10 , classified as ctts , also seems to be an interesting object .
a careful inspection of night to night observations of this star reveals amplitude variation .
its 2006 observations show that the brightness of this star was varying with larger amplitude ( 1.122 mag ) and period 1.102 day . from 2007 and
onwards , it shows consistent brightness variation with amplitude of @xmath30.20 mag and period of 1.102 day . whereas , the whole data set yield a period of 58.753 day .
star v19 ( ( @xmath43 mag ) could be a pms pulsating star as suggested by its variability characteristics like period , amplitude and the shape of light curve .
the period and amplitude of star v19 vary over time .
its period on 11 october 2009 and 27 october 2010 is found to be 0.101 day and 0.096 day , respectively , whereas its amplitude varied from 0.022 to 0.014 mag on respective dates .
considering all the observations of this star its period comes out be 0.109 day .
star v19 might be a @xmath25 scuti type variable star .
an early type star v23 ( brightest star in the present sample ) , considered to be a member of the cluster , might be a pulsating star .
its period ( 0.217 day ) and amplitude ( 0.027 mag ) are very much similar to the b - type pulsating stars .
star v21 , lying just below the ms , could be a cluster member .
the period , amplitude and the shape of light curve of v21 suggest that it might be a pulsating star .
the spectroscopic observations of star v21 was made by majaess et al .
they considered it to be a member of the cluster be 59 and found it to a be b - type star .
the field variables have period @xmath5 1.16 days and amplitude @xmath5 0.30 mag . in order to study the correlation between stellar activity and mass / age , we plot rotation period as a function of mass / age in fig
the rotation periods of ttss associated with the cluster be 59 do not show any dependency on the mass ( fig .
11 left panel ) .
fig.11 ( right panel ) shows two long periodic variables ( v3 with p=89.8 day and v10 with p=58.753 day , cf .
table 2 ) have ages less than 5 myrs , which could be interpreted as the rotation is regulated by disk locking at early ages ( jayawardhana et al .
2006 , herbst et al .
the presence of nir excess in @xmath38 of v3 and h@xmath21 emission in case of v10 is probably indicative of disk locking .
but it is worth to mention here that large ir excess alone could not indicate ongoing disk - locking because ir excess is the indicator for the presence of circumstellar dust and it is not an indicator for current mass accretion ( lamm et al .
2005 ) . fig .
12 ( left panel ) and fig .
12 ( right panel ) show amplitude as function of mass and age , respectively .
12 ( left panel ) reveals that the amplitude of ttss seems to be correlated with the mass of the star in the sense that amplitude decreases with increasing mass of the variables .
the largest amplitude is found in the case of v3 ( 0.855 mag ) and v10 ( 1.122 mag ) .
the star v3 shows significantly higher @xmath38 excess with relatively larger extinction , whereas star v10 shows h@xmath21 emission .
thus both the stars manifest presence of disk .
the decrease in amplitude could be due to the dispersal of the disk material .
if this notion is true , it indicates that the mechanism for disk dispersal operates less efficiently for relatively low mass stars .
in fact there is evidence in literature that disk dispersal mechanism depends on the mass of the stars ( e.g. carpenter et al .
but in fig 12 .
( right panel ) , the amplitude variation with ages shows no definite trend , and majority of the sources having the ages within 5 myrs show the amplitude of variability up to 0.35 mag with two cttss as exception , which have been discussed earlier .
the paper presents our efforts to search for variable stars in young open cluster be 59 . using @xmath44 photometry of a @xmath0 field around the cluster
we have identified 42 variable stars .
the probable members of the cluster are identified using @xmath1 colour - magnitude diagram and @xmath2 colour - colour diagram . out of 42 variables ,
31 variable stars are found to be the probable pms stars associated with the cluster be 59 . the majority of the probable pms candidate variable stars have ages 1 to 5 myrs .
the masses of probable pms variable stars range from @xmath30.3 to @xmath33.5 m@xmath20 , suggesting that these could be ttss .
the sample of ttss in the cluster be 59 indicates that @xmath390% ttss have periods less than 15 days .
the cttss have larger amplitudes in comparison to wttss
. the larger amplitude of cttss indicates the presence of hot spots on their surface .
the smaller amplitude in the case of wttss suggests that these pms stars might be in the process of dissipation of their circumstellar disk .
it is found that the amplitude of ttss decreases with increasing mass of the variable stars .
the decrease in amplitude of variability of ttss could be due to the dispersal of the disk material , indicating that the mechanism for disk dispersal operates less efficiently for relatively low mass stars .
authors are very thankful to the anonymous referee for useful suggestions , which improved the scientific contents of the paper .
alves de oliveira c. , casali m. , 2008 , a&a , 485 , 155 appenzeller i. , & mundt r. , 1989 , astron .
rev . 1 , 291 breger m. , 1972 , apj , 176 , 367 blanco v. m. , williams a. d. , 1959 , apj , 130 , 482 bessell m. s. , brett j. m. , 1988 , pasp , 100 , 1134 bouvier j. , malbet f. , monin j. -l . , 1994 , ap&ss , 212 , 159 cabrit s. , edwards s. , strom s. e. , strom k. m. , 1990 , apj , 354 , 687 calvet n. , hartmann l. , 1992 , apj , 386 , 239 carpenter j. m. , hillenbrand l. a. , skrutskie m. f. , 2001 , aj , 121 , 3160 carpenter j. m. , mamajek e. e. , hillenbrand l. a. , meyer m. r. , 2006 , apj , 651 , 49 catala c. , 2003 , ap&ss , 284 , 53 catala c. , donati j. f. , bhm t. , landstreet j. , henrichs h. f. , unruh y. , hao j. , collier c. a. , johns - krull c. m. , kaper l. , et al . , 1999 , a&a , 345 , 884 cutri r. m. , skrutskie m. f. , van dyk s. , et al . , 2003 , 2mass all sky catalog of point sources .
cohen j. g. , persson s. e. , elias j. h. , frogel j. a. , 1981 , apj , 249 , 481 chauhan n. , pandey a. k. , ogura k. , jose j. , ojha d. k. , samal m. r. , mito h. , 2011 , mnras , 415 , 1202 donati j. -f .
, semel m. , carter b. d. , rees d. e. , collier cameron a. , 1997 , mnras , 291 , 658 edwards s. , strom s. e. , hartigan p. , strom k. m. , hillenbrand l. a. , herbst w. , attridge j. , merrill k. m. , probst r. , gatley i. , 1993 , aj , 106 , 372 finkenzeller u. , mundt r. , 1984 , a&as , 55 , 109 grankin k. n. , melnikov s. yu . , bouvier j. , herbst w. , shevchenko v. s. , 2007 , a&a , 461 , 183 grankin k. n. , bouvier j. , herbst w. , melnikov s. y , 2008 , a&a , 479 , 827 girardi l. , bertelli g. , bressan a. , chiosi c. , groenewegen m. a. t. , marigo p. , salasnich b. , & weiss a. , 2002 , a&a , 391 , 195 herbst w. , herbst d. k. , grossman e. j. , weinstein d. , 1994 , aj , 108 , 1906 herbst w. , booth j. f. , korett f. l. , et al , 1987 , aj , 94 , 137 herbst w. , mundt r. , 2005 , apj , 633 , 967 herbst w. , bailer - jones c. a. l. , mundt r. , meisenheimer k. , wackermann r. , 2002 , a&a , 396 , 513 herbig g. h. , 1960 , apjs , 4 , 337 herbig g. , 1977 , apj .
214 , 747 haisch k. e. , lada e. a. , lada c. j. 2001 , apj , 553 , l153 hussain g. a. j. , 2002 , astron .
nachr . , 323 , 349 jayawardhana r. , coffey j. , scholz a. , brandeker a. , van kerkwijk m. h. , 2006 , apj , 648 , 1206 koenigl a. , 1991 , apj , 370 , 39 kurtz d. w. , marang f. , 1995 , mnras , 276 , 191 kundurthy p. , meyer m. r. , robberto m. , beckwith s. v. w. , herbst t. , 2006 , aj , 132 , 2469 lamm m. h. , mundt , r. , bailer - jones c. a. l. , herbst w. , 2005 , a&a , 430 , 1005 landolt a. u. , 1992 , aj , 104 , 340 lenz p. , breger , m. , 2005 , _ comm .
_ , 146,53 lomb n.r . , 1976 , apss , 39 , 447 lynden - bel d. , pringle j. e. , 1974 , mnras , 168 , 603 majaess d. j. , turner d. g. , lane d. j. , moncrieff k. e. , 2008 , javso , 36 , 90 marconi m. , ripepi v. , alcal j. m. , covino e. , palla f. , terranegra l. , 2000 , a&a , 355 , l35 messina s. , parihar p. , koo j .- r . , kim s .-
, rey s .- c .
, lee c .- u . , 2010 , a&a , 513 , 29 meyer m. r. , calvet n. , hillenbrand , l. a. , 1997 , aj , 114 , 288 nordhagen s. , herbst w. , rhode k. l. , williams e. c. , 2006 , aj , 132 , 1555 ojha d. k. et al .
, 2004a , apj , 608 , 792 ojha d. k. et al . , 2004b , apj , 616 , 1042 pandey a. k. , sharma s. , ogura k. , ojha d. k. , chen w. p. , bhatt b. c. , ghosh s. k. , 2008 , mnras , 383 , 1241 scargle j.d . , 1982 , apj , 263 , 835 scholz a. , eislffel j. , mundt r. , 2009 , mnras , 400 , 1548 shu f. h. , najita j. , ruden s. p. , lizano s. , 1994 , apj , 429 , 797 siess l. , dufour e. , forestini m. , 2000 , a&a , 358 , 593 strom k. m. , strom s. e. , edwards s. , cabrit s. , skrutskie , m. f. 1989 , aj , 97 , 1451 strom s. e. , strom k. m. , yost j. , et al , 1972 , apj , 173 , l65 .
skrutskie m. f. , dutkevitch d. , strom s. e. , edwards s. , strom k. m. , shure m. a. 1990 , aj , 99 , 1187 stetson p. b. , 1987 , pasp , 99 , 191 stetson p. b. , 1992 , j. r. astron .
can . , 86 , 71 van den ancker m. e. , de winter d. , tjin a djie h. r. e. , 1998 , a&a , 330 , 145 yang j. , fukui y. , 1992 , apj , 386 , 618 zwintz k. , marconi m. , reegen p. , et al .
2005 , mnras , 357 , 345 zwintz k. , weiss w. w. , 2006 , a&a , 457 , 237 zwintz k. , hareter m. , kuschnig r. , et al . , 2009 , a&a , 502 , 239 | we present time - series photometry of stars located in the extremely young open cluster berkeley 59 . using the 1.04 m telescope at aries , nainital ,
we have identified 42 variables in a field of @xmath0 around the cluster .
the probable members of the cluster are identified using @xmath1 colour - magnitude diagram and @xmath2 colour - colour diagram .
thirty one variables are found to be pre - main sequence stars associated with the cluster .
the ages and masses of pre - main sequence stars are derived from colour - magnitude diagram by fitting theoretical models to the observed data points .
the ages of the majority of the probable pre - main sequence variable candidates range from 1 to 5 myrs .
the masses of these pre - main sequence variable stars are found to be in the range of @xmath30.3 to @xmath33.5 @xmath4 and these could be t tauri stars .
the present statistics reveal that about 90% t tauri stars have periods @xmath5 15 days .
the classical t tauri stars are found to have larger amplitude in comparison to the weak line t tauri stars .
there is an indication that the amplitude decreases with increase of the mass , which could be due to the dispersal of disk of relatively massive stars .
[ firstpage ] [ firstpage ] open cluster : berkeley 59 colour magnitude diagram : variables - pre - main sequence stars |
on the exact solutions of integrable models , there is a new classification way recently based on the property of the spectral parameter @xcite : if the spectral parameter is positive , the related solution of the nonlinear system is called positon which is usually expressed by means of the triangle functions . if the spectral parameter is negtive , the related solution of the nonlinear system is called negaton which is usually expressed with help of the hyperbolic functions . the term
positon " and negaton " may be pursued back to matveev et al @xcite though the singular solutions had been studied earlier @xcite .
the so - called complexiton , which is expressed by combinations of the triangle functions and the hyperbolic functions , is related to the complex spectral parameters .
it is meaningful that the complexiton exists for _ real integrable systems .
the most of known positon solutions are singular . for various important integrable systems such as the kdv equation
there is no nonsingular positon found .
the negaton can be both singular and nonsingular .
especially nonsingular negatons are just the usual solitons .
the interaction solutions among positons and negatons have also been studied recently@xcite . unfortunately , except for a special c - integrable model , the sto ( sharma - tass - olver ) model @xcite , in our knowledge , there is no known nonsingular positon - negaton interaction solution . though the complexiton is proposed for several years , but up to now , one has not yet found a nonsingular complexiton solution for any ( 1 + 1)-dimensional integrable systems . in high dimensions , ( 2 + 1)- and ( 3 + 1)-dimensions ,
some special types of analytic nonsingular positon , posilton - negaton interaction solutions and the complexiton solutions can be easily obtained because of the existence of the arbitrary functions in their expressions of exact solutions@xcite . _
then an important and interesting question is naturally arised : + _ are there any nonsingular complexiton and positon - negaton solutions for s - integrable models ? as we all know , the korteweg - de vries ( kdv ) equation is a typical and classical model to describe the weakly shallow long waves in the soliton theory .
many important properties of kdv equation , such as infinitely many symmetries , infinitely many conservation laws @xcite , inverse scattering transform , bcklund transformation , darboux transformation , etc . have been studied extensively for a long time and it can be viewed as a completely integrable hamiltonian system @xcite .
nevertheless , it is known that both the positon and complexiton solutions of the kdv equation are singuler@xcite . _ on the other hand , coupled integrable systems , which come up in many physical fields , have also been studied extensively and a lot of interesting results are given out from both the classification view and application field @xcite .
the first coupled kdv system was put forward by hirota and satsuma ( hs ) in 1981 @xcite .
the multi - soliton solutions , conservation laws and darboux transformation , etc for the hs model are studied in detail @xcite .
many other coupled kdv systems such as the ito s system @xcite , the drinfeld and sokolov @xcite model , the fuchssteiner equation @xcite , the nutk - oguz model@xcite , the zharkov system @xcite and the foursov model @xcite are also constructed .
vladimir gives a classification of the integrable coupled kdv systems by using a symmetry approach @xcite . in the real application aspect of the coupled kdv systems , a quite general coupled kdv equation , @xmath0=0\label{ckdv1}\end{aligned}\ ] ] where @xmath1 and @xmath2 are arbitrary constants ,
is derived from a two - wave modes in a shallow stratified liquid @xcite .
another type of general coupled kdv system @xmath3 with arbitrary constants @xmath4 is obtained from a two layer model of the atmospheric dynamical system @xcite .
a more general coupled kdv system @xmath5 is derived to describe two - component bose - einstein condensates @xcite . in this letter
, we study the exact solutions , especially the positons , negatons and complexitons of a special coupled kdv system @xmath6 it is clear that the coupled kdv system is a special case for all the physical systems , and .
it is also interesting that , the system can be read off simply from the real and imaginary parts of the single complex kdv equation @xmath7 where @xmath8 from the fact that can be derived from the complex kdv equation , one can reasonable believe that all the integrable properties of the real kdv equation can be preserved .
actually , we have really proved its painlev property though we havent write down the proof procedure here because the method is standard and even can be performed simply by pressing the enter " key with help of some existent maple or mathematica programmes related to the painlev test , say , ptest " by xu and li @xcite .
in the remained part of this paper , we use the darboux transformation which is one of the most powerful tools in constructing exact solutions for many integrable nonlinear equations to the coupled kdv system to study its positon , negaton , complexiton solutions and the interaction property among these types of excitations .
in this section , we apply the darboux transformation to the coupled kdv system in order to construct exact solutions . because the coupled kdv system is obtained by imposing complex variables on the kdv equation , we can easily obtain the darboux transformation for the coupled kdv system from those of the kdv equation .
it is known that the lax pair for the kdv equation reads @xmath9 @xmath10 substituting @xmath11 into eqs . and , the lax pair for eq . is as follows : @xmath12 and @xmath13 then it is easy to prove that the compatibility conditions of four equations are exactly the equation system . with the help of the darboux transformation for the kdv equation ,
we can easily obtain the first step darboux transformation for the coupled kdv system @xmath14=u-{\ln(f^2+g^2)}_{xx},\ ] ] @xmath15=v-2{\arctan\left(\frac gf \right)}_{xx},\ ] ] and the new wave functions are transformed to @xmath16 @xmath17 where @xmath18 is a wave function solution of the lax pair with @xmath19 . in order to construct the second step darboux transformation ,
we need two wave function seed solutions @xmath20 and @xmath21 with two different spectral parameters @xmath22 and @xmath23 respectively , then the second step darboux transformation is @xmath24=u-\left[\ln(f^2+g^2)\right]_{xx},\ ] ] @xmath25=v-2\left[\arctan\left(\frac g f\right)\right]_{xx},\ ] ] where two functions @xmath26 and @xmath27 are given by @xmath28 and @xmath29 respectively with @xmath30 is the usual wronskian determinant .
similarly , the third and fourth step darboux transformations are written down as follows simply , @xmath31=u-[\ln(h^2+l^2)]_{xx},\ ] ] @xmath32=v-2\left[\arctan\left(\frac lh \right)\right]_{xx},\ ] ] where @xmath33 and @xmath34 with @xmath35 are solutions of the lax pairs with three different parameters @xmath36 and @xmath37 .
the fourth step darboux transformation is more complicated because more wave functions are included in .
the final results reads @xmath38=u-[\ln(p^2+q^2)]_{xx},\ ] ] @xmath39=v-2\left[\arctan\left(\frac qp \right)\right]_{xx},\ ] ] where @xmath40 and @xmath41 are @xmath42 and @xmath43 respectively , while @xmath44 are four wave function vector of the lax pair corresponding to four parameters @xmath45 in general , the @xmath46-step darboux transformation for the coupled kdv system is given by @xmath47=u-[\ln(w_r^2+w_i^2)]_{xx},\ ] ] @xmath48=v-2\left[\arctan\left(\frac { w_r}{w_i } \right)\right]_{xx}\ ] ] with @xmath49 and @xmath50 being the real and imaginary part of the wronskian of the @xmath46 complex wave functions @xmath51 i.e. , @xmath52
in this section , we systematically study the positon , negaton and complexiton solutions for the coupled kdv system by means of the darboux transformations given in the last section . based on the first step darboux transformation , analytical positon solutions for the coupled kdv system can be constructed directly .
taking the seed solution as @xmath54 , then solving the lax pair - directly , one can find @xmath55 @xmath56 substituting the result equations and into the first step darboux transformation yields the single general positon solution for the coupled kdv system : @xmath57}{(c_1 ^ 2\cos^2\xi_1+c_3 ^ 2\cos^2\xi_2)^2},\\ v&=&\frac{4c_1c_3\lambda\left[(c_1 ^
2\cos^2\xi_1-c_3 ^ 2\cos^2\xi_2)\cos(\delta_1-\delta_2 ) -(c_1 ^ 2-c_3 ^ 2)\cos\xi_1 \cos\xi_2\right]}{(c_1 ^ 2\cos^2\xi_1+c_3 ^ 2\cos^2\xi_2)^2}.\label{dv11}\end{aligned}\ ] ] it is easy to find that the positon solution is always nonsingular for @xmath58 . actually , from we know that the positon is singular only for @xmath59 while this constant condition leads to @xmath60 .
this result coincides with the fact that the positon of the real kdv is singular .
1 shows the nonsingular positon structure with the constant parameter selections @xmath61 at time @xmath62 .
epsf from fig . 1 , and the expressions and , we know that a nonsingular positon is just a special nonsingular periodic wave solution . in order to obtain the negaton solutions for coupled kdv system , we only need to substitute @xmath64 into eqs . and , then the wave functions are @xmath65 substituting the above wave functions into the expressions of the first step darboux transformation , we have the single negaton solution for the coupled kdv system @xmath66}{(c_1 ^ 2\sinh^2\eta_1+c_2 ^ 2\sinh^2\eta_2)^2},\\ v&=&\frac{2c_1c_2k^2\left[(c_1 ^ 2\sinh ( 2\eta_1)+c_2 ^ 2\sinh(2\eta_2)\right]\sinh(\delta_1-\delta_2 ) } { ( c_1 ^ 2\sinh^2\eta_1+c_2 ^ 2\sinh^2\eta_2)^2}.\label{sdv11}\end{aligned}\ ] ] fig .
2 shows a special negaton structure under the parameter selections @xmath67 at time @xmath62 . from the negaton expression
, one can see that the negaton solution of the model is nonsingular except that @xmath68 which is corresponding to the real kdv case , @xmath60 .
epsf with the n - step darboux transformations , one may obtain much richer solution structures including the interactions among different types of localized excitations .
a type of soliton solution including two negatons can be constructed easily from the two - step darboux transformation .
let two wave function vectors , @xmath69 and @xmath70 , are solutions of the lax pair with the seed @xmath71 and two spectral parameters @xmath72 respectively , then we have the two negaton solution expressed by with @xmath73 fig .
3 and fig.4 display the special two - negaton interaction procedure for the fields @xmath74 and @xmath75 described by with zero seed , wave function selections and the parameter selections @xmath76 at times @xmath77 and @xmath78 respectively .
epsf epsf in the same way , by selecting the spectral parameters appropriately in the n - step darboux transformations one can obtain the interaction solutions among positons and negatons .
for instance , based on the second step darboux transformation by choosing two different spectral parameters with opposite sign , @xmath79 , then we can obtain the positon - negaton interaction solutions for the coupled kdv system .
to show a concrete example , we further fix the spectral parameters as @xmath80 , @xmath81 and select the corresponding spectral functions in the forms @xmath82 @xmath83 @xmath84 @xmath85 substituting - and - into - with the initial seed solution @xmath86 , a special positon - negaton solution follows immediately .
fig . 6 and
7 show the interaction procedure expressed by - , - and the initial solution @xmath86 for the fields @xmath74 and @xmath75 respectively .
epsf epsf for many integrable system , complexiton solutions have been found @xcite . but it should be pointed out that the known complexiton solutions have singularities . it is lucky that the analytical complexiton solutions are found in this letter for the first time .
let two spectral parameters @xmath87 and @xmath88 be conjugated complex number , that is @xmath89 and @xmath90 , then we arrive at two coupled wave solutions with the initial solution @xmath86 , @xmath91 @xmath92 @xmath93 @xmath94 substituting - into - , we can obtain @xmath95\nonumber\\ & & + 4\lambda_1(c_4c_3-c_2c_1)\cosh[2\lambda_2(x+12t\lambda_1 ^ 2 - 4t\lambda_2 ^ 2)],\end{aligned}\ ] ] @xmath96\nonumber\\ & & + 4\lambda_1(c_2c_3+c_4c_1)\cosh[2\lambda_2(x+12t\lambda_1 ^ 2 - 4t\lambda_2 ^ 2)],\end{aligned}\ ] ] where @xmath97 are all constants . to show a detailed structure of the complexiton , we fix the constant parameters further .
the substitution of and into - with the fixing constants @xmath98 , yields the following analytical nonsingular complexiton solution @xmath99_{xx}=\frac{a(x , t)}{b(x , t)}\end{aligned}\ ] ] and @xmath100_{xx}=\frac{c(x , t)}{b(x , t)},\end{aligned}\ ] ] where @xmath101,\end{aligned}\ ] ] @xmath102 and @xmath103.\end{aligned}\ ] ] epsf epsf the complexiton solution displayed above is analytical without any singularity and it is the first time to find this type of analytical complexiton solutions for the integrable systems to our knowledge . fig . 7 and fig .
8 show the detailed structures of the fields @xmath74 and @xmath75 expressed by and respectively .
in this letter , exact solutions of a coupled kdv system which is obtained as a special case of the general coupled kdv systems derived from the two layer fluid dynamical systems and two - component bose - einstein condensates are studied in detail .
the model can also be simply read off from the well known single kdv equation by assuming its field is complex .
thanks to the model can be obtained from the complex kdv equation , the lax pair of the coupled kdv system follows immediately and so does the darboux transformation . to study the detailed structures of the localized and periodic excitations of the model , the first four step darboux transformations are explicitly given in terms of the real wronskian determinant forms while the general n - step darboux transformation is written down with help of the complex wronskian determinant form .
starting from the trivial seed solution , the first step of the darboux transformation leads to two types of exact solutions , the negaton and the positon which are related to negative and positive spectral parameters respectively .
though the single negaton solution for other integrable models , say , the single real kdv system , can be both singular and nonsingular , it can be only analytic for the coupled kdv system .
similarly , the positon solution can also be only analytic while for the single real kdv equation there is no analytic positon .
the solutions obtained from the second step darboux transformation and the trivial zero seed can be two negaton interaction solution , two positon solution , negaton - positon solution and a single complexiton dependent on the selections of the spectral parameters . the interaction among negatons is completely elastic . the positon - negaton solution can be considered as a single soliton solution with a periodic wave background .
the single real complexiton solution yields if two spectral parameters are complex conjugates .
the known complexiton solutions for other integrable systems are singular@xcite .
it is interesting that the complexiton solution obtained here for the coupled kdv system is nonsingular !
this type of nonsingular complexiton solution may exist for other types of coupled integrable systems . though the coupled kdv system studied here can be read off from the complex kdv equation , the solution properties are quite different .
for the real single component kdv equation , there is no nonsingular positons and complexitons .
however , for the coupled kdv system , negatons , positons and complexitons can be nonsingular . for the single kdv system ,
various other types of exact solutions , such as the rational solutions , cnoidal wave solutions , algebraic geometry solutions and @xmath104 function solutions have been obtained by many authors under different approaches .
the similar solutions for the coupled kdv system can be easily obtained in some similar ways .
however , the detailed properties of the solutions , especially there analytic behavior , must be quite different .
the more about the model and its possible applications in fluids , atmospherical dynamics and the bose - einstein condensates are worthy of studying further .
the authors are indebt to thanks dr . x. y. tang and prof .
y. chen for their helpful discussions .
the work was supported by the national natural science foundation of china ( no . 90203001 and no .
10475055 ) .
99 w. x. ma , phys .
lett . a 301 ( 2002 ) 35 ; w. x. ma and k. maruno , physica a * 343 * ( 2004 ) 219 . v. b. matveev and m. a. salle _ darboux transformations and solitons , springer - verlag 1991 ; v. b. matveev , phys .
a 166 ( 1992 ) 205 ; 209 ; 168 ( 1992 ) 463 ; j. math .
35 ( 1994 ) 2955 . c. rasinariu , u. sukhatme and a. khare , j. phys
gen , 29 ( 1996 ) 1803 .
m. d. kruskal , lect .
( 1974 ) 61 ; h. airault , h. p. mckean and j. moser , commun .
pure appl . math . 30
( 1977 ) 95 ; m. j. ablowitz and h. cornille , phys .
lett . a 72 ( 1979 ) 277
. z. j. lian and s. y. lou , nonlinear analysis , ( 2005 ) in press .
y. chen and q. wang , phys .
lett . a. ( 2005 ) in press .
s. y. lou , h. c. hu and x. y. tang , phys .
e. , 71 ( 2005 ) 036604 ; h. c. hu and s. y. lou , phys .
a 341 ( 2005 ) 422 .
s. y. lou , phys .
a 277 ( 2000 ) 94 ; x. y. tang , s. y. lou and y. zhang , phys .
e. , 66 ( 2002 ) 046601 ; h. c. hu , x. y. tang , s. y. lou and q. p. liu , chaos , solitons and fractals , 22 ( 2004 ) 327 .
r. m. miura , c. s. gardner and m. d. kruskal j. math .
* 9 * ( 1968 ) 1204 ; s. y. lou , j. math .
35 ( 1994 ) 2390 .
f. magri j. math .
* 19 * ( 1978 ) 1156 .
v. e. zakharov and l. d. faddeev , functional anal .
appl . * 5 * ( 1971 ) 280 .
v. v. sokolov and t. wolf , j. phys .
a : math . gen . * 34 * ( 2001 ) 11139 .
+ v. v. sokolov and s. i. svinolupov , theor .
* 100 * ( 1994 ) 959 .
r. hirota and j. satsuma , phys .
a * 85 * ( 1981 ) 407 .
r. hirota and j. satsuma , j. phys .
japan * 51 * ( 1982 ) 3390 .
+ r. dodd and a. p. fordy phys .
a * 89 * ( 1982 ) 168 .
+ h. c. hu and q. p. liu , chaos , solitons & fractals * 17 * ( 2003 ) 921 .
m. ito , phys .
lett . a * 91 * ( 1982 ) 335 .
v. g. drinfeld and v. v. sokolov , sov . j. math . *
30 * ( 1985 ) 1975 .
b. fuchssteiner , prog .
* 68 * ( 1982 ) 1082 .
y. nutku and o. oguz , nuovo cimento soc .
b * 105 * ( 1990 ) 1381 .
a. y. zharkov , j. symb . comput . * 15 * ( 1993 ) 85
. m. v. foursov , inverse problems * 16 * ( 2000 ) 259 .
j. a. gear and r. grimshaw , stud .
70 ( 1984 ) 235 ; j. a. gear , stud .
72 ( 1985 ) 95 .
s. y. lou , b. tong , h. c. hu and x. y. tang , coupled kdv equations derived from atmospherical dynamics , preprint arxiv : nlin.si/0508029 .
v. a. brazhnyi and v. v. konotop , phys .
e 72 ( 2005 ) 026616 .
g. q. xu and z. b. li , comp .
commun . 161
( 2004 ) 65 . | taking the coupled kdv system as a simple example , analytical and nonsingular complexiton solutions are firstly discovered in this letter for integrable systems .
additionally , the analytical and nonsingular positon - negaton interaction solutions are also firstly found for s - integrable model .
the new analytical positon , negaton and complexiton solutions of the coupled kdv system are given out both analytically and graphically by means of the iterative darboux transformations .
.2 in = 0.8 cm [ section ] [ section ] [ section ] [ section ] |
some of the most prominent emission lines in t tauri stars are from hydrogen . in the first few decades of t tauri research
the balmer emission lines were attributed to formation in energetic winds @xcite , but in the mid - nineties a reassessment of the hydrogen emission line profile morphology in the context of magnetospheric infall @xcite became the underpinning for the current paradigm that these young stars are in the final stages of accretion from a protoplanetary disk .
early successes of the model included generating balmer emission line profiles that are centrally peaked and have a small blue asymmetry , with inverse p cygni profiles at favorable inclinations @xcite .
a series of papers culminating in @xcite carried out a grid of radiative transfer models for line formation under the assumption of magnetospheric accretion in a dipole field geometry , finding good agreement in general profile morphology and line luminosity for hydrogen lines in many stars , including the infrared lines pa@xmath3 and br@xmath4 .
these models have more recently been incorporated into composite accretion and disk wind scenarios @xcite and into time dependent 3-d numerical simulations of magnetospheric accretion , including both dipole and multipole configurations , which may be aligned or misaligned with the stellar rotation axis @xcite , and accompanied by mhd winds from the inner disk @xcite . in general the magnetospheric accretion models
do a reasonable job of describing the morphology of the hydrogen line profiles and their luminosities , and are widely accepted as the origin of hydrogen emission in classical t tauri ( ctts ) stars .
however , shortcomings of these models have been identified by @xcite , @xcite and @xcite , based on comparing predictions and observations of line profile morphology , especially in the larger extent of the blue wing emission in observed compared to model profiles .
regardless of their origin , several groups have found well defined relations between the hydrogen line luminosity and the disk accretion rate , where the latter is assessed from accretion shock models of the excess optical and ultraviolet emission attributed to the post - shock heated photosphere at the magnetospheric footprints , that also includes contributions from the pre - shock gas @xcite .
this empirical correlation between hydrogen line luminosity and disk accretion rate , first established for pa@xmath3 and br@xmath4 by @xcite , has become a powerful means of determining disk accretion rates for ysos over a wide range of masses included embedded objects , where high extinction prevents optical and ultraviolet emission excess above the photosphere from being observed and modeled @xcite .
improved understanding of hydrogen emission in accreting stars will require a confrontation between the physical conditions required to produce the line luminosities and line ratios in combination with high resolution profile studies that define the kinematics of the line formation region .
to date , the hydrogen line luminosities in the magnetospheric accretion models are treated in an ad - hoc way , where accretion rates through the funnel flow set the density of infalling gas , and line luminosities require fine - tuning the temperature and size of the flow , with larger sizes and cooler temperatures in the higher accretion rate , higher line luminosity sources .
required temperatures range from lows of 6000
7000 k for high accretion rate sources to 10,000 20,000 k for low accretion rate sources @xcite .
attempts to self consistently assess heating and cooling in funnel flows are more restrictive in allowed temperatures @xcite and the inferred low temperatures can not account for the observed hydrogen line luminosities @xcite . however , this ad - hoc approach to determining the temperature in a funnel flow does result in a relation between line luminosity and accretion rate similar to what is observed by assessing accretion rates from excess continuum emission , providing an overall consistency with this interpretation of the hydrogen emission .
a more direct approach for assessing the physical conditions in the hydrogen line formation region is to use observed line luminosities and line ratios as direct diagnostics of the density and temperature based on calculations of atomic level populations .
natta , giovanardi , and palla took this approach in their 1988 paper , exploring the ionization and excitation structure of hydrogen lines formed in dense , cool winds from low luminosity pre - main sequence stars , with mass loss rates between @xmath5 and @xmath6 @xmath7 yr@xmath8 .
they found that line luminosities increase with mass loss rate , although different assumptions for the gas temperature and stellar radiation field produced a wide range of fluxes at a given mass loss rate .
they also pointed out that line ratios in the infrared may be able to discriminate among different models .
the advantage of using near infrared hydrogen lines as diagnostics of physical conditions is significant .
the line opacities are much smaller than the balmer lines , so they rarely show blueshifted absorption from a wind and have a lower frequency of redshifted absorption from infalling gas @xcite and extinction corrections are also smaller .
historically , the most widely examined ratio in near infrared lines is pa@xmath3@xmath9br@xmath4 .
@xcite found this ratio , coupled with the line luminosities , to be roughly consistent with the magnetospheric models described above for 19 ctts in tau - aur . in the magnetospheric accretion scenario
both lines are optically thick and pa@xmath3@xmath9br@xmath4 ratios are between 3 and 6 .
a study of ctts and brown dwarfs in @xmath10 oph found most of the ctts to have similar ratios to tau - aur but the accreting brown dwarfs and a few ctts showed pa@xmath3@xmath9br@xmath4 ratios @xmath11 2 , which were posited to result from low temperature , high optical depths spots in the shock heated photosphere @xcite .
a recent investigation of this ratio in 47 sources in the chameleon i and ii star formation region @xcite found that while many have ratios similar to those in tau - aur , almost half have ratios @xmath12 2 , again posited to form in very optically thick regions with @xmath13 k. a growing number of near infrared spectrographs permit a broad range of ratios from paschen and brackett series decrements to be determined simultaneously , which is crucial since the lines are quite variable .
a number of recent studies have compared these decrements to predictions of baker and menzel s case b for radiative ionization and recombination , which are available in an online database @xcite . in this recombination scenario
the near infrared lines are optically thin , the level populations are dominated by radiative cascade from the continuum , and collisional effects are included .
this approach has yielded some surprising results . in a study of 15 stars from tau - aur , @xcite found paschen and brackett decrements , taken as an average over all stars , to be best matched by case b conditions with @xmath14 k and
@xmath15 @xmath16 @xmath2 .
similar approaches for other individual stars indicate quite different conditions . for tw hya
, @xcite find series decrements matching case b for @xmath17 20,000 k and @xmath15 @xmath18 @xmath2 , while @xcite found reasonable agreement for case b predictions of @xmath17 10,000 k and @xmath15 between @xmath19 and @xmath20 @xmath2 from the brackett decrement for ru lup .
similar approaches have been taken for other ysos , where @xcite find decrements implying @xmath17 10,000 k and @xmath15
@xmath21 @xmath2 for the outbursting ctts ex lup and @xcite find @xmath22 10,000 k and @xmath15 = @xmath23 @xmath2 for the aebe star v921 sco .
this very diverse range of physical conditions inferred from comparing case b predictions to series decrements , with a range of over 6 orders of magnitude in density and temperatures from 1000 to 20,000 k , makes it questionable as to whether this scenario may be an appropriate choice for evaluating the physical conditions in the hydrogen line formation region in accreting young stars . spectacular new instrumentation , such as the vlt s x - shooter and crires ,
ensure that abundant near infrared spectra of star forming regions will be forthcoming and it is thus of interest to identify good diagnostics for interpreting these results .
a new set of local line excitation calculations by @xcite for physical conditions appropriate for winds and accretion flows in ctts offers an opportunity to evaluate hydrogen line ratios in series decrements and across series without the restrictions imposed by case b assumptions and over a wider range of physical conditions and atomic properties than in the earlier work of @xcite .
since these calculations , which evaluate line emissivities as a function of density , temperature , and ionizing flux , also include transitions of , , and , they offer the opportunity to evaluate physical conditions from many emission line ratios simultaneously , presenting more rigorous tests of the line formation region than previously available .
in this paper we look at various diagnostics from the paschen and brackett series for 16 t tauri stars and compare them to predictions for both case b and the kwan & fischer local line excitation models . the sample overlaps with those of @xcite and @xcite but the lines are resolved , allowing us to specify velocity limits in the paschen and brackett lines of each star that are not affected by redshifted absorption features , thus giving more precise line ratios of the emitting gas than lower resolution studies and also to make good assessments of the continuum emission excess .
the presentation includes
2 describing the sample and uncertainties in extinction ,
3 presenting the observed hydrogen line ratios and introducing several diagnostics to compare to model predictions ,
4 comparing the assumptions for case b and the kwan & fischer local line excitation calculations , and 5 comparing model predictions to observations and examining the role of extinction uncertainties .
we end with a discussion and conclusions in 6 and
our sample is 16 classical t tauri stars ( ctts ) observed with spex at the infrared telescope facility on 2006 november 26 and 27 .
it is largely drawn from the tau - aur star forming region , with spectral types from g8 to m2 , and selected to cover a broad range of emission excess and disk accretion rates .
the spectra were taken in the short - wavelength cross - dispersed ( sxd ) mode , with an 0.3@xmath24 by 15@xmath24 slit and a spectral extent from 0.8 to 2.4 @xmath25 m at a resolving power @xmath26 .
total exposure times ranged from 16 to 48 minutes for the program stars , with @xmath27 , yielding a continuum @xmath28 at @xmath29 , increasing to longer wavelengths .
the only unresolved binary in the sample is df tau , separation of 0.09 , of spectral types m2.0 and m2.5 with a flux ratio at @xmath30 of 1.62 @xcite .
these spectra were included in the study of fischer et al .
( 2011 ) , hereafter @xcite , and further properties of the sample are detailed there , along with specifics of the reduction and analysis of the spectra . in the 2011 paper
these spectra were used in conjunction with near simultaneous spectra from keck i s hires and keck ii s nirspec to derive continuum veiling and continuum emission excesses from 0.48 to 2.4 @xmath31 . here
we focus on the near infrared hydrogen emission lines in the spex spectra , which were shown by @xcite to have equivalent widths proportional to the excess continuum emission . as in previous studies ,
the _ veiling _ is defined as the ratio of excess to photospheric emission at a specific wavelength .
table [ t.sample ] identifies the sample , along with the literature spectral types used in fehk , the veiling at 1 @xmath25 m , @xmath32 , derived in fehk , and the emission equivalent widths for pa@xmath3 , pa@xmath4 , and br@xmath4 measured from the spex spectra .
the 1 @xmath25 m veiling ranges from just barely detectable ( 0.1 ) to quite significant ( 3.4 ) and the equivalent widths range from lows of a few tenths of an to as high as 28 for pa@xmath3 and 10 for br@xmath4 .
the error in equivalent width depends on the signal to noise in the continuum , the line to continuum ratio , and the width of the line ( several hundred km s@xmath8 ) .
most stars have equivalent width errors @xmath33% but for the three stars ( aa tau , bm and , lkca 8) with the smallest line / continuum ratios ( 1.02 to 1.05 at br@xmath4 ) errors can exceed 50% .
lccrrr aa tau & k7 & 0.2 & 0.3 & 0.8 & 1.2 + as 353a & k5 & 2.2 & 28 & 15 & 18 + bm and & g8 & 0.1 & 1.0 & 0.2 & 0.3 + bp tau & k7 & 0.4 & 9.6 & 6.5 & 4.4 + cw tau & k3 & 1.2 & 8.0 & 4.9 & 3.3 + cy tau & m1 & 0.2 & 0.9 & 1.0 & 1.0 + df tau & m2 & 0.5 & 5.0 & 3.8 & 3.3 + dg tau & k7 & 0.7 & 14 & 8.9 & 7.6 + dk tau & k7 & 0.6 & 4.0 & 4.2 & 2.5 + dl tau & k7 & 1.8 & 23 & 16 & 12 + do tau & m0 & 0.9 & 8.7 & 6.3 & 3.4 + dr tau & k7 & 3.4 & 24 & 18 & 8.9 + hn tau & k5 & 1.0 & 12 & 7.3 & 5.0 + lkca 8 & m0 & 0.2 & 0.3 & 1.0 & 1.0 + rw aur & k1 & 2.1 & 26 & 14 & 10 + uy aur & m0 & 0.6 & 3.4 & 3.9 & 2.0 lccccccc aa tau & 1.3 & 0.5 & 2.3 & 1.2 & & -1.99 & -8.87 + as 353a & 2.1 & 2.1 & 3.4 & 2.5 & 2.9 & 0.05 & -6.94 + bm and & 1.6 & 0.7 & 1.6 & 1.3 & & -1.25 & -8.61 + bp tau & 1.8 & 0.5 & 1.8 & 1.0 & 0.9 & -0.59 & -7.60 + cw tau & 2.1 & 2.0 & 2.4 & 2.4 & 3.0 & -0.74 & -8.00 + cy tau & 1.2 & 0.1 & 1.7 & 0.7 & 0.0 & -1.97 & -9.03 + df tau & 1.8 & 0.2 & 2.3 & 1.1 & 1.0 & -0.50 & -6.88 + dg tau & 3.9 & 1.0 & 3.9 & 2.4 & 3.5 & -0.15 & -7.15 + dk tau & 1.8 & 0.8 & 3.1 & 1.6 & 0.3 & -0.78 & -7.72 + dl tau & 3.0 & 1.4 & 3.0 & 2.1 & 1.4 & -0.34 & -7.20 + do tau & 3.0 & 2.0 & 3.5 & 3.0 & 1.5 & -0.62 & -7.37 + dr tau & 1.5 & 1.0 & 1.7 & 1.7 & 0.7 & -0.02 & -6.95 + hn tau & 3.1 & 0.4 & 3.1 & 1.3 & 3.0 & -1.11 & -8.34 + lkca 8 & 0.5 & 0.2 & 2.0 & 0.7 & & -2.37 & -9.33 + rw aur & 2.2 & 0.5 & 1.2 & 1.1 & 1.9 & 0.16 & -7.05 + uy aur & 1.5 & 0.6 & 3.1 & 1.5 & 0.3 & -0.87 & -7.59 luminosity , to the y band veiling , @xmath32 , from spex spectra of 16 stars . data points correspond to values based on the mean @xmath34 from the literature and error bars show how the quantities change over the full spread of @xmath34 reported for each star . [ f.acc ] ] our spex spectra were not calibrated on an absolute scale , but the shape of the continuum is well defined by the relative fluxes and we used the spex continuum flux adjacent to each line plus an estimate for extinction to restore equivalent width ratios to the line intensity ratios that are the basis for comparing to models .
the extinction is a much larger source of uncertainty .
table [ t.av ] illustrates the magnitude of this problem .
the @xmath34 in column 2 is from @xcite , found from fitting the observed spectral energy distribution from 0.8 to 2.4 @xmath31 with a combination of a main sequence spectral template with zero reddening plus a continuum excess anchored by simultaneously measured line veilings of photospheric features .
this technique , following the approach taken in @xcite for optical spectra , is in principle superior to deriving extinction from observed colors , since , as shown in @xcite , most t tauri stars have excess emission at all wavelengths , invalidating the standard technique of deriving @xmath34 from colors . however , as shown in @xcite , when applied in the near infrared , this technique yields @xmath34 that are often larger than those derived from optical studies , due at least in part to the use of main sequence spectral templates and the possible presence of large cool spots .
thus we also list in table [ t.av ] the minimum and maximum @xmath34 collected from the literature , the mean of these values ( including fehk ) , and a new estimate of @xmath34 that we will derive in this paper in section 5.3 , based on comparing observed line ratios to the local line excitation models of @xcite .
the sources for the literature values of @xmath34 are cited in the notes to table [ t.av ] .
we use the extinction law of @xcite with @xmath35 , as represented in the routine _ fm_unred.pro _ in the idl astronomy library , to convert between extinctions at different wavelengths .
also listed in table [ t.av ] are the accretion luminosities and mass accretion rates derived from the luminosity of pa@xmath3 .
the pa@xmath3 luminosities are derived from the line equivalent width , the absolute flux in the adjacent continuum , the extinction and the distance .
since we did not have absolute flux calibrated spectra we adopted continuum fluxes from the 2mass @xmath29 magnitudes and used distances of 140 pc , 200 pc and 440 pc for tau - aur @xcite , as 353a @xcite and bm and @xcite respectively . the accretion luminosity and mass accretion rates were calculated from pa@xmath3 luminosities following the calibration of @xcite and @xcite .
mass accretion rates also required estimates of the stellar mass and radii , made from application of the @xcite tracks to the effective temperatures and stellar luminosities from @xcite . for extinctions we used the mean value of @xmath34 identified in table [ t.av ] .
both accretion luminosity and mass accretion rate are plotted in figure [ f.acc ] against the simultaneously observed 1 @xmath25 m veiling , @xmath32 .
the symbols correspond to the values derived with the mean @xmath34 while the vertical lines extending below / above each symbol correspond to values derived from the minimum / maximum values of @xmath34 from the literature cited in table [ t.av ] .
the figure illustrates that even with the large spread in reported @xmath34 , @xmath32 is a reasonable proxy for accretion luminosity and disk accretion rate , where the accretion luminosities for our sample range from 1.5 @xmath36 to @xmath37 @xmath36 and the disk accretion rates from @xmath38 to @xmath39 @xmath7 yr@xmath8 .
we will adopt the @xmath34 from fehk in calculating emission intensity ratios since it is determined in the same way for all stars in our sample .
a difference of 1 magnitude in @xmath34 affects the ratios pa@xmath4@xmath9pa@xmath3 and br@xmath40@xmath9br@xmath4 by a factor of 7% , br@xmath4@xmath9pa@xmath3 by 12% , and pa12@xmath9pa@xmath3 by 19% .
there is no correlation between line ratios and @xmath34 , indicating there is not a systematic effect on the ratios from the adopted extinction .
an examination of the effect of extinction in the comparison of observed line ratios to those predicted from line excitation models will be the focus of section 5.3 .
in this section we present line intensity ratios for the paschen series pa@xmath3 through pa12 , and for two brackett lines , br@xmath4 and br@xmath40 .
we will compare the 1 @xmath25 m veiling , @xmath32 , to observed line ratios and make use of the fact that the broad paschen and brackett lines are resolved with the modest spex velocity resolution of about 150 km s@xmath8 .
we calculate intensity ratios for pa@xmath4/pa@xmath3 and for br@xmath4/pa@xmath3 in three different ways for each star in our sample .
one ( ` direct ' ) is the intensity ratio based on the emission equivalent width measured directly from the spex spectra over the full range of the emission .
the other two are measured from _ residual line profiles _ , where the photosphere of an appropriately veiled spectral template is subtracted from the ctts spectrum ( see @xcite for a description of this technique and fehk for the templates used here ) . in one case
the equivalent widths are measured over the full range of the residual emission , and in the other the equivalent widths , and corresponding line ratios , are found over a limited velocity range in the residual line profile .
the line ratios from each of these three techniques are tabulated in table [ t.intensity ] for each star .
the residual pa@xmath3 , pa@xmath4 , and br@xmath4 profiles ( normalized and superposed ) are shown in figure [ f.vel ] with stars arranged in order of @xmath32 , from highest to lowest . although the spex resolution of 150 km s@xmath8 just resolves these broad lines , it is sufficient to show the opacity dependent differences in the redshifted absorption in some stars ( e.g. cw tau , dk tau , df tau , bm and ) .
we therefore compute the intensity ratios from residual profiles both over the full range of emission ( labeled ` full ' ) and also over velocity intervals selected to be free of opacity effects from the redshifted absorption ( labeled ` select ' ) , in principle giving more reliable ratios when comparing to models .
the velocity intervals are marked in figure [ f.vel ] with vertical dotted lines and identified in table [ t.intensity ] .
intensity ratios determined from emission equivalent widths in these velocity intervals are given for pa@xmath4@xmath9pa@xmath3 , br@xmath4@xmath9pa@xmath3 and br@xmath40/br@xmath4 in table [ t.intensity ] and the paschen series through pa12 , relative to pa@xmath3 , in table [ t.decrement ] .
we omit higher series lines because the lines are weak in most stars and the @xcite calculations used a 15-level h atom , giving reliable fluxes only through pa13 and br13 .
lcccccclll aa tau & 2.23 & 0.55 : & 0.86 & 0.49 : & 0 & 150 & 0.81 & 0.22 : & + as 353a & 0.47 & 0.69 & 0.69 & 0.45 & -400 & 200 & 0.68 & 0.27 & 0.82 + bm and & 0.16 & 2.20 : & 0.21 & 3.25 : & -300 & -200 & 0.46 & 0.17 : & + bp tau & 0.57 & 0.36 & 0.64 & 0.32 & -200 & 200 & 0.62 & 0.17 & 0.40 + cw tau & 0.44 & 0.71 & 0.61 & 0.70 & -300 & 0 & 0.76 & 0.34 & 0.80 + cy tau & 1.11 & 0.39 & 0.69 & 0.33 & -150 & 150 & 0.68 & 0.12 & + df tau & 0.66 & 0.47 & 0.76 & 0.35 & -150 & 0 & 0.66 & 0.19 & 0.57 + dg tau & 0.36 & 0.93 & 0.78 & 0.45 & -200 & 200 & 0.76 & 0.25 & 0.68 + dk tau & 0.68 & 0.51 & 0.83 & 0.33 & -300 & 0 & 0.73 & 0.18 & + dl tau & 0.49 & 0.59 & 0.81 & 0.34 & -400 & 300 & 0.85 & 0.22 & 0.77 + do tau & 0.45 & 0.61 & 0.68 & 0.36 & -150 & 150 & 0.67 & 0.16 & + dr tau & 0.63 & 0.44 & 0.80 & 0.35 & -300 & 300 & 0.80 & 0.24 & 0.70 + hn tau & 0.39 & 0.85 & 0.70 & 0.41 & -200 & 200 & 0.68 & 0.22 & 0.37 + lkca 8 & 2.64 & 0.54 : & 1.01 & 0.51 : & -150 & 0 & 0.86 & 0.37 : & + rw aur & 0.46 & 0.50 & 0.71 & 0.36 & -300 & 300 & 0.70 & 0.22 & 0.39 + uy aur & 0.58 & 0.41 & 0.69 & 0.35 & -200 & -200 & 0.64 & 0.16 & 0.19 lccccccc as 353a & 0.68 & 0.62 & 0.54 & 0.55 & 0.51 & 0.42 & 0.41 + bp tau & 0.62 & 0.44 & 0.39 & 0.34 & 0.22 & 0.14 & 0.13 + cw tau & 0.76 & 0.63 & 0.57 & 0.63 & 0.51 & 0.45 & 0.41 + df tau & 0.66 & 0.45 & 0.48 & 0.31 & 0.07 & 0.12 & 0.12 + dg tau & 0.76 & 0.67 & 0.64 & 0.55 & 0.48 & 0.46 & 0.37 + dk tau & 0.73 & 0.49 & 0.50 & 0.47 & 0.16 & 0.13 & 0.11 + dl tau & 0.85 & 0.72 & 0.72 & 0.74 & 0.64 & 0.55 & 0.46 + do tau & 0.67 & 0.48 & 0.45 & 0.34 & 0.30 & 0.11 & 0.10 + dr tau & 0.80 & 0.71 & 0.59 & 0.54 & 0.48 & 0.42 & 0.35 + hn tau & 0.68 & 0.59 & 0.46 & 0.36 & 0.31 & 0.23 & 0.27 + rw aur & 0.70 & 0.64 & 0.48 & 0.24 & 0.34 & 0.27 & 0.28 + uy aur & 0.64 & 0.49 & 0.42 & 0.22 & 0.22 & 0.17 & 0.11 for most of our sample the difference between direct and residual profiles is @xmath4110% .
however for the four stars with the smallest emission equivalent width and lowest @xmath42-band veiling ( @xmath43 , aa tau , bm and , cy tau and lkca 8) , these two approaches yield significant differences in profile morphology and equivalent width , as illustrated in figure [ f.residual ] for pa@xmath3 .
three of these four stars show residual br@xmath4 profiles in figure [ f.vel ] that are poorly defined , and their small line to continuum ratios and equivalent widths yield equivalent width errors in the selected velocity intervals @xmath44 50% .
we flag these large errors in the table and exclude these three stars in the subsequent figures that include the ratio of br@xmath4@xmath9pa@xmath3 .
one other star , cy tau , also shows a noisy residual br@xmath4 profile with a small br@xmath4 equivalent width in the selected velocity interval with an error around 30% .
although we include this star in the figures , the error in its br@xmath4@xmath9pa@xmath3 ratio is significant in comparison to the other stars , and as will be seen in later sections it is an outlier in some relationships .
we note that in contrast to previous papers that focus only on pa@xmath3 and br@xmath4 , here we put pa@xmath3 in the denominator of our ratios rather than br@xmath4 .
this facilitates comparisons among a wider range of lines , and is also more meaningful in comparison to the model predictions discussed in the next section . for direct ( solid ) and residual ( dash - dot ) profiles , along with the wtts template ( dotted ) for the four stars with smallest pa@xmath3 equivalent width in the direct spectra . in these four stars
the underlying photospheric absorption affects the observed profile and the residual profiles have different equivalent width and kinematic structure . for the remaining 12 stars the difference between direct and residual profiles is @xmath12 10% .
[ f.residual ] ] @xmath9pa@xmath3 line intensity ratios for the 16 stars in our spex sample based on three assessments of the emission equivalent width ( ew ) .
upper panel : ew measured directly from the original spectrum over the full range of velocity .
middle panel : ew measured from the residual profile over the full range range of velocity .
bottom panel : ew measured from the residual profile , over selected velocity intervals shown in figure 2 .
the spread in ratios and the mean value are reduced when selected velocity intervals from the residual profiles are used .
the colors correspond to three levels of @xmath42-band veiling ; high is red with @xmath45 , medium is gray with @xmath46 and low is blue : @xmath47 .
[ f.hist ] ] we show in figure [ f.hist ] a comparison of the pa@xmath4@xmath9pa@xmath3 emission intensity ratios for the 16 sample stars from the three approaches ( direct full , residual full , residual selected ) in histogram form , where veiling groups identified as high ( @xmath48 ) , medium ( @xmath49 ) , and low ( @xmath47 ) are separately colored .
the figure shows that the dispersion in the intensity ratio is largest for the ` direct ' method and smallest for the ` selected ' method , suggesting that the range of actual line ratios may be smaller than would be inferred from the standard approach .
we will adopt the narrower range of ratios from selected velocity intervals in residual profiles for the analysis in the next section in the interest of making the best comparison to model predictions .
we identify three relations in figure [ f.relations ] that will be the basis for comparing observations to the theoretical predictions from both case b and the @xcite , local line excitation calculations .
these are the paschen decrement normalized to pa@xmath3 , for pa@xmath4 through pa12 , plus two ratio - ratio relations , one between pa@xmath4@xmath9pa@xmath3 and br@xmath4@xmath9pa@xmath3 , and the other between br@xmath40@xmath9br@xmath4 and br@xmath4@xmath9pa@xmath3 , again sorted into high , medium and low @xmath42-band veiling groups .
@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 ( 14 stars ) , lower : ratio - ratio relation for br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 ( 11 stars ) .
color indicates levels of @xmath42-band veiling : red for @xmath45 , gray for @xmath46 and blue for @xmath47 .
there is a tendency for stars with higher @xmath32 to have larger line ratios .
[ f.relations ] ] because three low veiling objects in our sample of 16 ctts have unreliable br@xmath4@xmath9pa@xmath3 ratios , four have unreliable paschen decrements , and six have unreliable br@xmath40@xmath9br@xmath4 ratios we introduce in this and all subsequent figures spex - based line ratios for one additional low veiling ctts taken from the literature .
this is tw hya , an object with @xmath50 , using line intensity ratios taken from @xcite .
we have a number of @xmath42-band nirspec spectra for this object taken between 2006 to 2011 and consistently find the veiling to be very low or zero @xcite , and adding it improves the statistics for the low veiling group .
( although our nirspec profiles of tw hya show a weak redshifted absorption at pa@xmath4 , the ratios from vacca and sandell correspond to the ` direct full ' ratio method defined here . )
we see in all three relations that there is a spread in the observed ratios with a tendency for stars with higher @xmath32 to have higher ratios than stars with lower @xmath32 .
however , with such a small sample it is not clear how robust this result is .
we note that two of the low veiling stars from our sample that do not appear in figure [ f.relations ] due to their highly uncertain br@xmath4@xmath9pa@xmath3 ratios ( aa tau and lkca 8) have pa@xmath4@xmath9pa@xmath3 ratios at the high end of the observed values , suggesting that the trend we see here with veiling needs to be tested further .
we also note that the slight rise in the paschen decrement at pa9 is an artifact due to blending from adjacent emission lines in the stars with higher veiling and stronger line emission .
hydrogen line ratios depend on physical conditions in the line formation region .
baker and menzel s case b for radiative ionization and recombination has been successfully applied to understanding ionized nebulae such as hii regions and planetary nebulae for many decades and has recently been invoked to infer physical conditions in t tauri systems from line ratios of the paschen and brackett series of hydrogen ( see introduction ) , making use of the interactive online server that calculates hydrogen line ratios for a range of electron densities and temperatures @xcite .
the results , primarily focusing on the behavior of series decrements , suggest a surprisingly diverse range of implied electron densities and temperatures in the hydrogen line formation region that calls into question the applicability of the case b assumptions for t tauri stars .
the recent local line excitation calculations of @xcite , hereafter kf , developed to interpret t tauri spectral lines , offer an alternative option for inferring physical conditions from observed hydrogen line ratios . here
we review both approaches , and in the next section compare predicted ratios to observed values in both scenarios .
the line diagnostics we will explore in each scenario arise from a similar range of electron density and temperature , however the physical conditions are actually quite different . the case b calculations output line ratios as a function of input electron density @xmath15 and temperature @xmath51 .
no restrictions on @xmath15 and @xmath51 are explicitly imposed by the case b online simulator @xcite . by definition
, however , the recombination model needs the population of the @xmath52 level to be sufficiently small that collisional excitation from that level as a means of balmer , paschen , brackett , etc .
line emission is not significant .
this stipulates that the neutral hydrogen column density , @xmath53 ( where @xmath54 is the emission length scale ) , or equivalently , the ly@xmath55 optical depth , be sufficiently small that radiative de - excitation of @xmath52 occurs more rapidly than collisional excitation from @xmath52 .
since in case b @xmath15 is a free parameter , the ratio @xmath56 is not specified or determined , nor is the photoionization rate from the ground state , @xmath57 .
thus the recombination model implicitly assumes that @xmath57 and @xmath54 have values that ensure sufficiently small line optical depths .
the kf calculations are more general in including both recombination and collisional excitation as a means of producing line photons and in exploring the full range of line optical depths .
atomic parameters include 15 distinct energy levels of hydrogen , 19 of and key transitions of , and . upon inputs of the local physical conditions of hydrogen nucleon number density @xmath0 , temperature @xmath51 , ionization rate @xmath57 ( photoionization rates from excited states
are also included based on an assumed stellar plus veiling continuum ) , and the velocity gradient @xmath58 ( giving the emission length scale in a differentially moving medium ) , the calculations solve for the ionization fraction @xmath59 , level population ( including @xmath60 ) , and all line optical depths self - consistently .
the full set of input parameters , with @xmath51 from 5000 to 30,000 k , @xmath0 from @xmath61 to @xmath62 @xmath2 , @xmath57 of @xmath63 or @xmath64 @xmath65 , and @xmath58 of 150 km s@xmath8/2@xmath66 or 150 km s@xmath8/1.25 @xmath66 , presented in kf are designed to approximate conditions in the region of a wind or an accretion flow where the bulk velocity is @xmath11 150 km s@xmath8 .
the variations of the line emissivity ratios arising from the choices of @xmath57 and @xmath58 are very much smaller than those arising from density and temperature changes , and the results shown here are from the case of @xmath57 = @xmath63 s@xmath8 and @xmath58 = 150 km s@xmath8/2@xmath66 .
although each line emissivity is for a uniform density and temperature in roughly the middle of an accretion flow / wind , while the observed line flux will be an integration of the emissivity over the entire kinematic structure , kf pointed out that since each position in the flow will be represented by the local line excitation , only with different parameters , that it is possible to judge how the resultant ratios will be affected when averaged over a range of density and/or temperature .
they concluded that observed line flux ratios do indeed indicate clearly enough the physical conditions and that this approximation is reasonable .
the two models have similar ranges in @xmath15 , but refer to quite different physical regimes , arising from the fundamental difference in the energy source for the line photons , which is continuum photons more energetic than 13.6 ev in the recombination model and thermal kinetic energy in the kf model . the derived electron fraction in the kf calculations
is somewhat dependent on the ionization rate , as shown in their figure 5 , but @xmath67 is @xmath68 over a wide range of density for @xmath69 8750 k and @xmath11 0.1 at @xmath70 k. thus although @xmath15 in both the case b and kf models is similar , the line optical depths are vastly different . in kf , at fixed values of @xmath57 , @xmath58 , and @xmath51 , the run of the calculation with increasing @xmath0 corresponds to increasing line optical depths , producing a corresponding variation of the line emissivities . for example , the ly@xmath55 optical depth for the case of @xmath71 @xmath2 and @xmath22 10,000 k is @xmath72 .
this is much higher than the limit imposed by the recombination model , since in kf collisional excitation from @xmath52 , whose population is sustained by the strong ly@xmath55 trapping , is the predominant cause for the strength of the line emission .
consequently , the behavior of hydrogen line ratios with density and temperature is quite different in the two models .
this is illustrated in figure [ f.theory ] where the relation between increasing @xmath15 for case b or @xmath0 for the kf local line excitation is shown for four ratios : pa@xmath4@xmath9pa@xmath3 , br@xmath4@xmath9pa@xmath3 , pa12@xmath9pa@xmath3 , and br10@xmath9br@xmath4 for a range of temperatures .
for the kf models temperatures are shown from 5000 k to 20,000 k but for case b we include temperatures as low as 1000 k in order to reach the full range of the observed ratios .
the dispersion in observed values from tables [ t.intensity ] and [ t.decrement ] is indicated by a vertical line along the left side of each panel .
the observed ratios cover about a factor of two for pa@xmath4@xmath9pa@xmath3 and br@xmath4@xmath9pa@xmath3 and about a factor of four for pa12@xmath9pa@xmath3 and br10@xmath9br@xmath4 . for case b ( left , @xmath73 20,000 k ) and @xmath0 for kf ( right , @xmath74 20,000 k ) .
in case b , the line ratios have a very small dynamic range for @xmath75 k. in the kf calculations , the line ratios are initially low when the lines are optically thin and then increase steeply as @xmath0 increases , until they plateau at high optical depths .
the range of observed ratios is shown along the left side of each panel .
temperature line types apply to all panels .
[ f.theory ] ] for the kf models , the four ratios have the same general behavior with density and temperature , transitioning near @xmath76 @xmath2 from low optically thin values to higher ones , climbing steeply as the density and line optical depth increase .
( see figures in kf for corresponding optical depths ) .
the rise begins at lower densities for higher temperatures , but the general behavior is similar for temperatures between 5000 and 20,000 k. these relations readily account for the range of observed ratios , where higher ratios are explained by higher densities / optical depths . at the highest density in the kf calculations of @xmath77 @xmath2 ,
the ratios of pa@xmath4@xmath9pa@xmath3 and pa12@xmath9pa@xmath3 exceed unity and are still climbing , while the ratios of br@xmath4@xmath9pa@xmath3 and br10@xmath9br@xmath4 have plateaued at peak values around 0.6 and 2 , respectively .
in contrast , the case b predictions show quite different behavior . at low @xmath15
all four ratios are higher when @xmath51 is lower .
this is because the radiative recombination rate to level @xmath78 is proportional to @xmath79 , where the index @xmath55 increases from 0.7 to 1.1 as @xmath78 increases from 3 to 12 , so as @xmath51 decreases there is a stronger preference , comparatively , towards radiative recombinations into higher @xmath78 levels .
as @xmath15 increases , three - body recombinations begin to contribute . at a given @xmath51 these recombinations , unlike radiative ones , strongly favor population into higher @xmath78 levels , since the rate to level @xmath78 is proportional to @xmath80 times a factor which itself increases rapidly with increasing @xmath78 .
this rate is also roughly proportional to @xmath81 , so its influence is also stronger when @xmath51 is lower .
hence the dispersion of each depicted line ratio with temperature increases with increasing @xmath15 until collisional de - excitation and collisional ionization begin to dominate over radiative decay and the ratio falls with further @xmath15 increase .
this transition occurs at lower @xmath15 for higher @xmath78 levels because collisional rates are higher while einstein @xmath82 rates are lower for higher @xmath78 .
while it occurs near @xmath83 @xmath2 for pa@xmath4@xmath9pa@xmath3 and br@xmath4@xmath9pa@xmath3 , in the case of pa12@xmath9pa@xmath3 it occurs at @xmath84 @xmath2 .
( the density grid in the storey and hummer online server can interpolate ratios only between integer values of log @xmath15 , so whether the peak ratio is precisely at the stated value is unclear ) .
note that the case b relations require temperatures @xmath11 1000 k to account for ratios above the median observed value for each ratio shown , while electron densities @xmath85 @xmath2 are favored for pa@xmath4@xmath9pa@xmath3 and br@xmath4@xmath9pa@xmath3 but @xmath15 @xmath41 @xmath86 @xmath2 is required for ratios incorporating the higher levels pa12@xmath9pa@xmath3 and br10@xmath9br@xmath4 .
thus , the differing behaviors of the line ratios with increasing @xmath15 shown in figure [ f.theory ] result from the fundamentally different nature of the two calculations . in the kf calculations the behavior follows simply from the fact that as @xmath0 increases at a local point with all the other parameters fixed , the line optical depths also increase .
the rise of each ratio with @xmath0 follows from the stronger build - up of population into higher @xmath78 levels via collisions , coupled with the larger optical depth of the lower transition , thus enhancing the emissivity of the line in the numerator and reducing the emission effectiveness of the line in the denominator . in the recombination model
, there is no explicit determination of @xmath87 , but given that @xmath87@xmath88 ) , where @xmath89 and @xmath90 are the radiative and three - body recombination coefficient respectively , @xmath91 is the proton density and neglecting , for ease of explanation , the process of collisional ionization from high @xmath78 levels , @xmath87 will , at fixed @xmath57 , increase rapidly as @xmath15 increases , so to maintain the same constraint on @xmath53 in order for recombination to remain dominant over collisional excitation from @xmath78 = 2 as a means of photon production , @xmath57 must increase or @xmath92 must decrease accordingly .
indeed , with collisional rates increasing with increasing @xmath15 while einstein @xmath82 rates remain fixed , the limit on @xmath53 actually needs to vary inversely with @xmath15 , making the required changes in @xmath57 and/or @xmath92 even more drastic .
thus the recombination results actually implicate associated changes in those two parameters which one must also assess for plausibility before applying the case b results to observations .
for example , if the length scale of the emission corresponding to the same velocity gradient used in kf is adopted , the resultant constraint on the ly@xmath55 optical depth of @xmath93 requires @xmath94 @xmath2 for the case b conditions .
then , for @xmath95 5000 k , @xmath96 ( @xmath97 s@xmath8 . for @xmath98 @xmath2 , for example , the photoionization rate from the ground state
must then be orders of magnitude higher than the @xmath57= @xmath63 s@xmath8 considered in the kf calculations .
( the fairly high ionization in kf with @xmath59 of @xmath44 0.6 for @xmath99 k is due to photoionization from excited states by the stellar and veiling continuum . )
the length scale of the emission region can be reduced to alleviate the constraint on @xmath57 , but probably not by more than a factor of 10 .
thus in case b it is the requirement that recombinations continue to predominate at high @xmath15 , demanding a physical regime requiring particularly strong photoionizations to maintain low line optical depths , that is the reason for the depicted behavior of the line ratios with @xmath15 in figure [ f.theory ] .
these restrictions embedded within the case b assumptions are grounds to doubt their validity in the interpretation of hydrogen line ratios in ctts . to further demonstrate this point , in the next section we compare conclusions of both case b and the kf calculations to hydrogen line ratios .
here we return to the three diagnostics shown in figure [ f.relations ] . for both case
b and the kf local line excitation calculations , we compare observational diagnostics to theoretical predictions for ( 1 ) the paschen decrement normalized to pa@xmath3 , for pa@xmath4 through pa12 , ( 2 ) the ratio - ratio relation of pa@xmath4@xmath9pa@xmath3 and br@xmath4@xmath9pa@xmath3 , and ( 3 ) the ratio - ratio relation of br@xmath40@xmath9br@xmath4 and br@xmath4@xmath9pa@xmath3 .
the comparison between case b predictions and observations is shown figure [ f.b_nh ] for the paschen decrement and in figure [ f.b_ratios ] for the two ratio - ratio relations .
we have clarified in the previous section the rather stringent requirements that allow case b to be applied as @xmath100 increases but continue with the comparison since this model has been widely applied to observed ctts line ratios . in order to keep the figure from being too busy
we restricted the range of @xmath15 ( in units of @xmath2 ) to four orders of magnitude , adopting the range most often cited in comparison to observed ratios , for log @xmath101 , 10 , 11 , 12 .
the predicted paschen decrement in figure [ f.b_nh ] is shown in four panels , one for each electron density , each with iso - temperature surfaces from 1000 to 20,000 k. the predicted decrements differ primarily in the maximum value of pa@xmath4@xmath9pa@xmath3 ( highest for @xmath102 ) and in the rate of decline from lower to higher paschen lines ( shallowest for @xmath103 and @xmath104 k ) . the higher veiling stars , with a combination of a high pa@xmath4@xmath9pa@xmath3 ratio and a shallow decline down the series , are not well matched by case b. the lower veiling stars are more readily accounted for , with small pa@xmath4@xmath9pa@xmath3 and a steep decrement to pa12 for a range of electron densities and temperatures from 10,000 to 20,000 k. ) .
each panel features a different log @xmath15 ( 9 , 10 , 11 , 12 ) , for a range of temperatures from 1000 to 20,000 k. [ f.b_nh ] ] and @xmath51 compared to observations . _
upper _ : theoretical relations for pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 are shown in different line types for temperatures between 1000 k and 20,000 k. iso - density lines connect ratios for @xmath105 , 10 , 11 , 12 , 13 ( @xmath106 , 10 are degenerate ) .
_ lower _ : similar relations for br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 ( @xmath107 is out of the plot range ) . in both panels observations
are colored according to @xmath32 group .
[ f.b_ratios ] ] the iso - density lines behave differently in each ratio - ratio plot in figure [ f.b_ratios ] due to the different behavior of line ratios in case b with increasing @xmath15 seen in figure [ f.theory ] . in the pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 relation ,
the locus of iso - density lines rotates counter - clockwise around the figure , with the widest range of allowed ratios for the lowest temperatures ( @xmath108 k ) .
in contrast in the br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 relation the locus of iso - density lines rotates clockwise around the figure , and again the widest range of allowed ratios is for the lowest temperatures ( @xmath108 k ) .
the observational trend for a pair of ratios to increase together is largely a temperature effect in case b , where the highest ratios require temperatures below 3000 k. however , an uncertainty of @xmath34 of only one magnitude is sufficient to move an individual star through a wide range of @xmath15 and temperature loci , putting stringent , and probably unrealistic , requirements on the precision of @xmath34 in order to make a reliable comparison . with that caveat ,
most observed points in the pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 relation suggest log @xmath15 from 1012 and temperatures over the full range from 1000 to 20,000 k. these results can be compared with previous studies comparing line ratios to case b. the conclusion that the best case b fits to series decrements required @xmath103 and @xmath109 k by @xcite was based on averaging line ratios for a group of tau - aur stars with high veiling . while this conclusion is obtained under case b to match the shallow fall - off of the series decrement in the high veiling stars ( but not the low veiling stars ) , the observed pa@xmath4@xmath9pa@xmath3 is too high in comparison to the case b predictions for these conditions .
the case b fit for paschen and brackett series decrement found by @xcite for the low @xmath42-band veiling star tw hya , with @xmath110 and @xmath17 20,000 k , is consistent with its comparison with all three diagnostics here .
however most ctts would not be compatible with these findings for tw hya . in sum ,
case b predictions are questionable for ctts .
the problem is most severe for the high veiling stars which are not well described by the predicted paschen decrement while the ratio - ratio plots suggest a wide range of electron densities and temperatures among the stars .
the low veiling stars are reasonably well matched to predicted paschen decrements for @xmath111 and @xmath112 10,000 k , although in the ratio - ratio plots , except for tw hya , they require electron densities at least an order of magnitude lower and a range of temperatures .
this is consistent with the diverse range of temperatures and electron densities reported in the literature for classical t tauri stars when being compared to case b predictions . ) .
each panel features a different @xmath0 ( 0.25 , 1 , 1.6 , 2.5 @xmath113 ) , for a range of temperatures from 7500 to 12,500 k.[f.kf_nh ] ] @xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 are shown in different line types for temperatures between 5000 k and 20,000 k. iso - density lines connect ratios for @xmath114 , 11 , 11.3 . _
lower _ : similar relations for br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 . in both panels observations
are colored according to @xmath32 group .
[ f.kf_ratios ] ] the comparison between the kwan & fischer ( kf ) local line excitation predictions and observations is shown in figure [ f.kf_nh ] for the paschen decrement and in figure [ f.kf_ratios ] for the two ratio - ratio relations . predicted decrements are again shown in four panels , but in this case only for a total range of a factor of 10 in @xmath0 in units of @xmath2 , from 0.25 to @xmath115 , with iso - temperature surfaces from 7500 to 12,500 k. this limited density range covers all of the observed ratios for the cases with @xmath99 k. in this temperature range the model curves are very similar , in that at a fixed temperature the pa@xmath78/pa@xmath3 value is small for each @xmath78 when the density is low , and the whole decrement curve rises as density increases , with a higher rise at higher @xmath78 .
the rise of the decrement to a flatter shape with increasing density happens more slowly when @xmath116 k , so that for @xmath51 between 5000 and 7500 k ( not shown in the figure ) the required range of densities to cover all the observations is somewhat higher , from @xmath117 to @xmath118 @xmath2 .
the same range of densities is shown in the ratio - ratio plots , which extend over a wider range of temperature from 5000 k to 20,000 k. the behavior of each ratio - ratio relation is similar , with line ratios increasing with increasing density and higher temperatures corresponding to a higher value of pa@xmath4@xmath9pa@xmath3 or br@xmath40@xmath9br@xmath4 for a given value of br@xmath4@xmath9pa@xmath3 . in this figure
the iso - density contours bend back to somewhat lower ratios for @xmath116 k , again reflecting the need for somewhat higher densities at lower temperatures to reproduce a given ratio .
several things are apparent when comparing the kf relations to the observed values . in the series
decrement plots in figure [ f.kf_nh ] , there is a well defined trend such that as pa@xmath4@xmath9pa@xmath3 increases , so does pa@xmath78/pa@xmath3 , in agreement with the observations .
this behavior is not seen in the case b decrements in figure [ f.b_nh ] . in the ratio - ratio plots in figure [ f.kf_ratios ] , the upward slope toward higher ratios in both pairs of lines
is aligned with the predicted behavior for line ratios increasing with increasing density , covering a span from @xmath11 @xmath119 to @xmath120 @xmath2 for @xmath121 8750 k , where stars with higher accretion rates have densities at the higher end of this range .
moreover , in contrast to case b , here the same conclusions are reached from all three diagnostic relations .
all indicate the density in the hydrogen line formation region spans an order of magnitude and increases with increasing mass accretion rate .
temperatures are not well constrained , since , as seen in the ratio - ratio plots , reddening vectors cut the closely spaced iso - temperature lines almost orthogonally and errors in @xmath34 of @xmath122 magnitudes span the full range of possible temperatures .
however the densities are robustly determined and are not sensitive to extinction uncertainties .
thus , in contrast to case b , the kf local line excitation calculations give consistent results across all diagnostics for the density , and suggest densities in the hydrogen line formation region are similar among the ctts in our sample , on the order of @xmath123 .
there is also a suggestion that higher veiling stars on average have densities higher than low veiling stars by about a factor of @xmath124 .
the temperature is not as readily identified , since uncertainties in extinction result in ratios that can intersect any of the temperature contours between 5000 and 20,000 k. in contrast to the much lower temperatures inferred from comparisons of line ratios with case b @xcite , this range corresponds to temperatures where collisional excitation as a means of photon production is important .
the previous figures applied the @xmath34 from fehk to convert equivalent width ratios to intensity ratios .
however , the range in @xmath34 reported in the literature is considerable so in the upper panel of figure [ f.ext ] we again show the pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 relation comparing observations to the kf predictions , but this time show the ` observed ' intensity ratio implied for each assessment of @xmath34 from the literature cited in table [ t.av ] , with from 2 to 7 @xmath34 per star .
for comparison , we also include the directly observed ratio with no correction for extinction .
in contrast to the minimal impact of the @xmath34 spread ( up to 2.5 magnitudes for individual stars ) on the relation between the accretion luminosity and the veiling shown in figure [ f.acc ] , in a ratio - ratio plot the @xmath34 spread moves the ` observed ' ratios through a significant domain of model predictions in the direction of a reddening vector .
most apparent from the figure is that the extrema of the reported @xmath34 for each star often lie outside the range expected from either the kf or case b relations for the range of densities and temperatures explored here .
the fehk values used here are also higher than the average value for most stars , however the largest values typically come from the study of @xcite , while the smallest values are from @xcite and @xcite .
@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 and lower is pa@xmath4@xmath9pa@xmath125 versus br@xmath4@xmath9pa@xmath125 . in both panels
dotted lines connect ratios from the minimum and maximum reported @xmath34 for each star , where the minimum ( @xmath126 ) is the ratio uncorrected for extinction .
other points correspond to : @xmath34 from @xcite ( @xmath127 ) , @xmath34 s from the literature ( @xmath128 ) , and the kf @xmath34 ( @xmath129 ) . in both cases ,
extrema of the literature values are outside the range of the model .
both use the same temperature legend , but a wider range of densities is shown in the lower panel . [ f.ext ] ] even with the uncertainty in @xmath34 some conclusions can be drawn from comparing observed ratios to the kf calculations .
first , the uncertainty in extinction does not alter the empirical conclusion that stars with higher veiling tend to have higher line ratios , since this trend is not affected by reddening vectors . in the kf models , the reddening vectors cross the iso - temperature lines orthogonally so that @xmath0 but not temperature is well constrained , with an implied range in density from @xmath119 to @xmath120 @xmath2 , as noted in the last section . the temperature is impossible to disentangle from the uncertainty in @xmath34 ; however , if we make the assumption that ctts hydrogen lines form in a narrow temperature range then the spread in observed ratios across iso - temperature lines in the ratio - ratio plots could be due wholly to uncertainty in the adopted extinction for each star . adopting this assumption with the kf predictions
we can recover an @xmath34 that would place each star along a single temperature locus .
while no such temperature assessment exists from first principles , @xmath130 10,000 k is a reasonable choice for both the magnetospheric accretion models of @xcite and @xcite and the wind models of @xcite . we thus calculate a new @xmath34 for each star , indicated by filled triangles in figure [ f.ext ] and listed in column 6 of table [ t.av ] , corresponding to the extinction correction required to bring each star to the @xmath22 10,000 k iso - temperature line from the kf models . this ` kf @xmath34 ' is in the mid - range of previously determined values for most stars , and is on average 0.7 magnitudes smaller than the @xmath34 adopted throughout this paper from the fehk study . for tw hya the ` kf @xmath34 ' is zero since it already lies on the @xmath22 10,000 k iso - temperature line .
this is also the extinction that is found in all studies of this nearby star .
this technique can not be applied to cy tau , since the uncorrected ratios lie on the @xmath22 20,000 k iso - temperature line .
thus although we identify ` kf @xmath34 ' = 0 for this star , this is not in line with previous estimates and likely means the uncorrected ratios are in error due to the large uncertainty in the definition of the br@xmath4 profile .
in contrast if the same spread of implied ratios were plotted on the case b ratio - ratio plot of figure 8 the result would be ambiguous in inferring both density and temperature .
the spread of up to 2.5 magnitudes in @xmath34 for each star means that reddening vectors would cross multiple iso - density lines orthogonally , and each iso - temperature surface would also be intersected twice , corresponding to a different implied density in each case . under this scenario , even if one were to adjust the @xmath34 to locate all the stars along the same iso - density line , temperatures from 1000 to 20,000 k would be required to match the observations , and tw hya would be several orders of magnitude higher density than the other stars .
it is difficult to imagine that there is such a variety of physical conditions in the region where hydrogen lines form in ctts , making the kf assumptions seem far more plausible than any case b scenario
. additional insight on extinction can be gleaned by turning to line ratios that arise from the same upper level , so the effects of density and temperature on the ratio are much reduced . to this end
, we show in the lower panel of figure [ f.ext ] a ratio - ratio plot of pa@xmath125@xmath9pa@xmath4 versus br@xmath4@xmath9pa@xmath125 .
the behavior of the kf calculations for this ratio - ratio relationship is different from those shown previously because both br@xmath4 and pa@xmath125 arise from the same upper level ( @xmath131 ) .
this ratio is shown as a function of @xmath0 and temperature in the kf calculations in figure [ f.pdel ] , analogous to the ratios shown in figure [ f.theory ] .
however , this ratio behaves quite differently from the other line ratios studied here , which rise continuously as the optical depth increases until reaching a plateau value .
instead , br@xmath4@xmath9pa@xmath125 departs from the optically thin ratio around @xmath0 = @xmath86 @xmath2 , first rising as pa@xmath125 becomes optically thick , and then falling as the optical depths of both lines continue to increase .
when included in a ratio - ratio plot , such as pa@xmath125@xmath9pa@xmath4 versus br@xmath4@xmath9pa@xmath125 , the former ratio traces density while the latter is sensitive mostly to extinction , with a dynamic range about half that of the other line ratios over the same span of density and temperature .
@xmath9pa@xmath125 as a function of log @xmath0 .
since both transitions arise from the same @xmath131 upper level , even with a wide range of optical depths , this ratio is much less sensitive to density and has a smaller dynamic range than other ratios shown in fig .
[ f.pdel ] ] returning the to lower panel of figure [ f.ext ] , we have included the corresponding set of ` observed ' ratios for the same literature values of @xmath34 shown in the upper panel .
again , the largest values of literature @xmath34 are outside the range of the models and the high veiling stars require an @xmath34 in excess of unity . for two thirds of the stars the ratios resulting from the ` kf @xmath34 ' , set by the @xmath22 10,000 k iso - temperature locus in the ratio - ratio plot of pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 in figure [ f.ext ] , fall on the same locus here suggesting it may offer a reasonable estimate for the extinction in many cases .
clearly there is a need for a more definitive assessment of extinction in t tauri stars but that is not the purpose of this work .
we do however conclude that the highest values of @xmath34 reported in the literature seem to be out of bounds for many stars .
similarly some of the low extinctions in the literature are incompatible with either the case b or kf model for the five high mass accretion rate stars with strong emission excess at all wavelengths noted by fehk ( cw tau , dg tau , as 353a , hn tau and rw aur ) . for these stars , a@xmath132 in excess of 1 magnitude is the minimum required to move the line ratios into the realm of the models . while we do not consider our ` kf @xmath34 ' values to be definitive , since the assumption that the temperature in the hydrogen line formation region is identical in all ctts is an oversimplification , it will be of interest to compare them to new assessments of @xmath34 that will be possible with the new generation of high resolution and broad wavelength coverage spectrographs .
correcting for extinction is notoriously difficult for embedded class i protostars , when the @xmath133 and @xmath30 bands may be the shortest accessible wavelengths and scattered light from envelopes as well as emission from accretion related phenomena are present .
often a variety of techniques to recover @xmath34 are employed , and the result for a single star can differ by many magnitudes @xcite .
the near infrared ratio - ratio relations investigated here offer an alternate way of addressing this problem . while the locus of theoretical predictions in the ratio - ratio planes are similar in both case b and the kf calculations ,
we have shown in previous sections that those from kf provide a more likely scenario for the observed line ratios and we thus focus this discussion on the kf predictions .
if an observed pair of hydrogen line ratios , such as pa@xmath4/pa@xmath3 versus br@xmath4/pa@xmath3 , or for more embedded objects br10/br@xmath4 versus br@xmath4/pa@xmath3 , are tracked back along a reddening vector until they intersect the kf model relations , an @xmath34 can be recovered if a temperature is adopted . however , since the range of iso - temperature loci from 5000 k to 20,000 k span only two magnitudes along a reddening vector , any adopted temperature in that range will give @xmath34 with a fractional uncertainty that is modest for deeply embedded sources . as an illustration , in the top panel of figure [ f.ext2 ]
we show the ratio - ratio relation of pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 for the kf calculations along with observations for the ten objects in l1641 for which intensities for all three lines were reported in a recent study by @xcite .
the study focussed on a mix of class i and ii sources observed with ntt sofi .
we show the observed ratios for these ten objects , both uncorrected for extinction , and corrected with the extinction adopted by the authors , assessed from the mean of up to six different @xmath34 estimates after rejecting outlying values .
if the reddening vectors are extended back from the uncorrected ratios to the @xmath22 10,000 k line in the kf relations , they would extend well beyond the adopted ratios in half of the objects , suggesting that the adopted @xmath34 , which range from 2 to 10 , may be significantly underestimated in some sources .
@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 for 10 sources in l 1641 @xcite , plus the kf relations . both directly observed ( @xmath126 ) and
extinction corrected ( @xmath134 ) values are shown , using extinctions adopted by the authors .
lower : ratio - ratio plot of br10@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 for 12 deeply embedded class i sources in orion and the kf relations .
reddening vectors extend from the directly observed values to the kf @xmath22 10,000 k locus . in both panels ,
the density increases along the model tracks from lower left to upper right , as in figure 10 , except over a wider range from @xmath135 to @xmath136 @xmath2 .
[ f.ext2 ] ] as a second illustration , we show in the lower panel of figure [ f.ext2 ] the ratio - ratio relation of br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 for the kf calculations along with observations of 12 deeply embedded protostars in orion observed with spex as part of the hops open - time key program of the herschel space observatory @xcite .
the uncorrected line ratios for these protostars , a subset of class i sources with @xmath137 mag identified from 2mass and spitzer colors between 1.2 and 24 by @xcite , show that the extinctions are considerably higher than the objects from caratti o garrati .
( there are no objects in common in the two samples . ) using the technique of extending reddening vectors from the uncorrected ratios to the @xmath22 10,000 k line in the kf relations , the ` kf @xmath34 ' fall between 10 and 34 magnitudes .
the seds for these objects are being analyzed by the hops team , who determine extinctions by comparing the 1 to 870 spectra and photometry of each object to a grid of radiative transfer models first described by @xcite and generated with the code of @xcite , which yields the extinction from the intervening interstellar dust , the circumstellar envelope , and the accretion disk . as will be shown in a forthcoming paper ( fischer et al .
, in prep . ) , the extinction derived in these two approaches is comparable , suggesting both may be a viable means of determining extinction to embedded sources , in contrast to commonly used approaches for class i extinctions . in these illustrations not only
can an @xmath34 be recovered from the intersection of the reddening vectors with the loci of the kf iso - temperature calculations , but this intersection also implies a unique value of the hydrogen number density , @xmath0 . for the 10 sources in l1641 the implied densities fall in the range @xmath138 @xmath2 , somewhat higher than for the ctts studied here . for the 12 sources from across the orion complex ,
the implied densities span a wider range , from @xmath139 @xmath2 .
while further investigation of the validity of the kf models applied to accreting systems is required , we consider this a promising approach to further our understanding of class i sources as well as ctts .
paschen and brackett lines are some of the strongest lines in near infrared spectra of accreting young stars
. they are superior to the balmer series for line ratio diagnostics since they are less subject to uncertainties in extinction , and to the opacity effects that riddle the balmer series with blueshifted and redshifted absorption .
we have focussed here on velocity resolved paschen and brackett lines in a small sample of classical t tauri stars , primarily in tau - aur , in order to look for variations in the hydrogen line ratios as a function of mass accretion rate and to compare them to theoretical models for hydrogen line formation .
the primary strength of this work is to show the limitations of using case b predictions to interpret ctts hydrogen line ratios and to demonstrate the potential of using the @xcite local line excitation predictions to infer physical conditions and extinction .
we experimented with three methods for determining hydrogen line intensity ratios from emission equivalent widths . for the majority of our sample , those with strong emission and no redshifted absorption , we found almost identical ratios among the three methods .
however , when the emission equivalent width is weak and/or if redshifted absorption is present , the line ratios corrected for these effects can be altered from directly observed values .
for example in three stars the directly observed ratio of pa@xmath4@xmath9pa@xmath3 exceeded unity but dropped to @xmath140 after correcting the directly measured line profile for the underlying veiled photospheric absorption and taking ratios only in regions of the line profile free of redshifted absorption .
while the ratio of pa@xmath4@xmath9pa@xmath3 can exceed unity in the kf models at densities exceeding @xmath0 = 10@xmath141 @xmath2 , the ratio of br@xmath4@xmath9pa@xmath3 is always below unity .
thus it is surprising that while half the stars in a low resolution study of chameleon i and ii were found by @xcite to have br@xmath4@xmath9pa@xmath3 ratios similar to objects in tau - aur ( 0.1 to 0.4 ) , the remainder had larger ratios , with br@xmath4@xmath9pa@xmath3 between 0.5 and 2 .
such large ratios are inconsistent with any case b scenario and outside of the range of kf calculations as well . as noted by the authors , these ratios could be explained by invoking line formation in optically thick ( i.e. , blackbody ) lte emission with @xmath104 k. however , the br@xmath4 intensity could appear to exceed pa@xmath3 at low resolution if the lines are subject to redshifted absorption , which would reduce emission from pa@xmath3 relative to br@xmath4 , and might go unnoticed in unresolved lines .
high resolution spectra of the chamaeleon objects are required to see whether this group of stars has near infrared line profiles that differ from all other ctts observed to date , which are generally similar in their kinematic properties and thus suggestive of formation under similar conditions @xcite . after correcting the intensity ratios for the effects mentioned above
, we found a loose relation between near infrared hydrogen line ratios and @xmath42-band veiling , and thus the implied mass accretion rate . in the kf predictions , which we favor over case
b , this would correspond to somewhat higher densities in the hydrogen line formation region in stars with higher accretion rates . in a previous study we found that pa@xmath3 line profiles from nirspec at @xmath142 25,000 showed kinematic behavior that correlated with the @xmath42-band veiling , where stars with higher veiling had broader line profiles @xcite .
taken together , these results suggest stars with higher disk accretion rates have both higher densities and higher velocities in the line formation region .
these high veiling stars also show extended blue wings with velocities in their hydrogen profiles in excess of what can be produced in magnetospheric infall , so winds along with funnel flows are likely implicated .
a larger sample is required to see whether this connection between veiling and line ratios is real or is simply due to our small sample size , and whether there is a connection to the profile morphology .
both case b and kf local line excitations predict a locus of line ratios that overlap with observations of ctts ; however , the behavior of the line ratio diagnostics leads us to favor the kf predictions over those of case b. the inconsistencies in interpreting line ratios among different case b diagnostics likely arises because the condition that the neutral hydrogen column density , @xmath53 , is sufficiently small that radiative de - excitation occurs more rapidly than collisional excitation from @xmath52 , required to keep the paschen and brackett lines optically thin , is violated .
additional evidence for the latter is the fact that 24% of ctts show redshifted absorption in pa@xmath4 when examined at high spectral resolution @xcite , which requires line opacities well in excess of unity . in contrast
the kf predictions cover a large range of possible line opacities , self consistently taking into account the local density , temperature , and ionization rate .
to enable others to compare their data with these calculations , we have assembled a set of hydrogen line ratios from the kf local line excitation models and make them publicly available on a web server for other researchers acquiring near infrared spectra of ctts .
the site includes both the ratios used here , plus others based on h@xmath55 , br@xmath55 ( 4.05 @xmath25 m ) , h76 ( 12.4 @xmath25 m ) and h97 ( 11.3 @xmath25 m ) that may be useful for existing or planned observations . under the kf assumptions ,
the implied density in the hydrogen line formation region is within a factor of a few times @xmath117 @xmath2 for the 16 ctts in our sample , with the higher accretion rate stars at the higher end of the range .
here we compare these densities to those expected from hydrogen lines that are formed in accretion funnels , and find they are somewhat lower than expected based on the models of @xcite , also adopted by @xcite . in these models
balmer , paschen and brackett emission arises over the full length of the accretion columns and the mass accretion rate sets the density in the accretion columns . for example , in figure 2 of @xcite , for a fiducial case of an aligned , symmetric dipole flow with a maximum temperature of 8000 k and @xmath143 = @xmath144 @xmath7 yr@xmath8 , the density increases from @xmath118 to @xmath145 @xmath2 along the region of the accretion column where infall velocities exceed 100 km s@xmath8 and much of the line emission arises .
this @xmath143 is in line with about half the stars in our sample , based on pa@xmath3 line luminosities and shown in figure 1 , yet the hydrogen number densities we infer for these stars are at least an order of magnitude lower . for the remainder of our sample ,
mass accretion rates are between @xmath146 and @xmath5 @xmath7 yr@xmath8 , which would have correspondingly lower densities in the funnel flow model .
thus although the general behavior expected in the funnel flow models , that higher accretion rate objects would have higher densities in the accretion columns , is born out in the observations , there are discrepancies . whether these inconsistencies are serious challenges to the common assumption that hydrogen lines are formed primarily in funnel flows or not is difficult to say at this stage of the investigation . another approach is to note that in magnetospheric accretion models the density in the immediate pre - shock gas is about an order of magnitude higher than that over most of the flow due to the channeling effect of the ` funnel ' ( e.g. , figures in @xcite and @xcite ) .
thus if we amplify the number density we infer in the hydrogen line formation region , @xmath147 @xmath2 , by an order of magnitude to @xmath148 @xmath2 to reflect the corresponding immediate pre - shock densities , we can compare this to pre - shock densities based on modeling either optical / uv continuum emission excess or line ratios in the x - ray domain .
for example , @xcite define an energy flux @xmath149 carried into the accretion shock by the funnel flow , where @xmath150 . in their models
the free fall velocity @xmath151 is kept constant at around 300 km s@xmath8 and the density @xmath152 in the immediate pre - shock gas sets the accretion shock energy . the correspondence between @xmath143 and @xmath153 is given by their equation 11 .
applying this with the same @xmath143 as in the fiducial case cited above , @xmath154
@xmath7 yr@xmath8 , the corresponding @xmath155 erg @xmath156 s@xmath8 implies a pre - shock number density @xmath18 @xmath2 , again about an order of magnitude higher than inferred from the near infrared line ratios . for stars at the low end of the mass accretion rates thought to apply to our sample , @xmath157 @xmath7 yr@xmath8 , the pre - shock number density would be two orders of magnitude smaller at @xmath117 @xmath2 , in this case larger than we would infer for the pre - shock density of these stars
. we can also look at the pre - shock densities derived from modeling the he - like triplets of and in x - ray spectra .
these line ratios in ctts can not be explained by coronal emission and are attributed to formation in the accretion shock @xcite .
a study of these line ratios in several low accretion rate ctts indicate pre - shock densities @xmath18 @xmath2 for two stars in our current study , bp tau and tw hya @xcite , again an order of magnitude higher than we would infer . moreover , although these pre - shock densities are in line with those from magnetospheric accretion models , the inferred mass accretion rates based on x - ray line ratios are @xmath158 @xmath7 yr@xmath8 for bp tau and @xmath159 @xmath7 yr@xmath8 for tw hya @xcite , an order of magnitude or more lower than would be inferred from figure 1 based on pa@xmath3 luminosities or from the accretion shock models of @xcite based on optical / uv continuum excess .
the tendency for mass accretion rates inferred from x - ray line ratios to be consistently lower than those based on accretion shock models of optical / uv emission excess is recognized by @xcite , who suggest that this might arise from inhomogeneous spots , partial absorption in buried shocks , or the presence of accretion streams that impact at velocities considerably below free - fall speeds ( see also @xcite ) . in this paper
we have applied the kf calculations only to hydrogen line ratios .
the kf local line excitation calculations also include , , , and so that when ratios of other species are included , we will be able not only to test the conclusions regarding density , but also set constraints on the temperature as well .
although the accretion funnel paradigm as a source of the near infrared hydrogen emission lines has withstood more than a decade of scrutiny , there are growing indications that a deeper investigation is warranted . in an era where it will now be possible to simultaneously match line luminosities , line ratios , and high resolution line profiles from multiple lines over a wide spectral range and over several stellar rotation periods to those predicted from modern magnetospheric accretion models @xcite , with misaligned fields with multipole components , coupled with rigorous diagnostics of the physical conditions in the line formation region
, we may be able to decipher what the relative contributions of funnel flows , accretion shocks , winds and the inner disk might be to the rich emission line spectra of ctts .
the key conclusions from this work are * in our limited sample , we see a tendency for the hydrogen line ratios in ctts to have different behavior in stars of high and low veiling . in the higher veiling stars the paschen decrement is shallower and line ratios tend to be higher in ratio - ratio relations such as pa@xmath4@xmath9pa@xmath3 versus br@xmath4@xmath9pa@xmath3 and br@xmath40@xmath9br@xmath4 versus br@xmath4@xmath9pa@xmath3 .
this conclusion is independent of extinction uncertainties .
however it is based on a relatively small sample of stars and needs to be examined in a larger context .
* the @xcite local line excitation calculations offer a more consistent interpretation of physical conditions in the hydrogen line formation region of t tauri stars than case b. under kf assumptions the density in the hydrogen line formation region lies within the range @xmath160 @xmath2 , with densities higher on average in stars with higher accretion rates .
extinction uncertainties preclude a reliable temperature determination . under case
b assumptions not only do different diagnostics yield different implied values of temperature and electron density in ctts , the range of implied values exceeds four orders of magnitude in electron density and a factor of ten in temperature . *
the largest source of uncertainty in determining line ratios in t tauri stars is in correcting for extinction .
the range of @xmath34 values in the literature can be 2.5 magnitudes for a given star .
this makes comparisons with line excitation models problematic for typical t tauri stars , but for deeply embedded sources , where an error in @xmath34 of two magnitudes is less consequential , the kf models can be used to evaluate both the extinction and the density in the line formation region when near infrared hydrogen line ratios can be determined .
thanks to greg herczeg , mike peterson , and hans moritz gnther for useful discussions . also , thanks to our referee who made some very helpful comments .
the authors reverently acknowledge the cultural significance of the mauna kea summit to the indigenous hawaiian community .
we are most fortunate to have had the opportunity to conduct observations with irtf from this mountain . | in accreting young stars one of the prominent spectral features in the near infrared is the paschen and brackett series in emission .
we examine hydrogen line ratios for 16 classical t tauri stars from spex spectra and and assess the trends with veiling and accretion .
the observed line ratios are compared to two theoretical models for line formation : ( 1 ) baker and menzel s ( 1938 ) case b for radiative ionization and recombination and ( 2 ) a set of local line excitation calculations designed to replicate the conditions in t tauri winds and magnetic accretion columns @xcite . while the comparison between case b and
observed line ratios implies a wide range in electron density and temperature among the hydrogen line formation regions in t tauri stars , the predictions of the local line excitation models give consistent results across multiple diagnostics . under the assumptions of the local line excitation calculations ,
we find that @xmath0 in the hydrogen line formation region is constrained to @xmath1 @xmath2 , where stars with higher accretion rates have densities at the higher end of this range . because of uncertainties in extinction ,
temperature is not well delineated but falls within the range expected for collisional excitation to produce the line photons .
we introduce new diagnostics for assessing extinction based on near infrared hydrogen line ratios from the local line excitation calculations . |
two points inside a polygon @xmath3 are _ visible _ to each other if their connecting segment remains completely inside @xmath3 .
the _ visibility polygon _ ( @xmath7 ) of a point @xmath8 inside @xmath3 ( or @xmath9 ) is the set of vertices of @xmath3 that are visible from @xmath8 .
there have been many studies on computing @xmath7 s in simple polygons . in a simple polygon @xmath3 with @xmath10 vertices ,
@xmath9 can be reported in time @xmath11 by spending @xmath12 time and @xmath13 of preprocessing space @xcite .
this result was later improved by @xcite where the preprocessing time and space were reduced to @xmath4 and @xmath5 respectively , at the expense of more query time of @xmath14 .
the visibility problem has also been considered for line segments .
a point @xmath15 is said to be _ weakly visible _ from a line segment @xmath1 if there exists a point @xmath16 such that @xmath17 and @xmath15 are visible to each other .
the problem of computing the _ weak visibility polygon _ of @xmath1 ( or @xmath2 ) inside @xmath3 is to compute all points of @xmath3 that are weakly visible from @xmath1 .
if @xmath3 is simple ( with no holes ) , chazelle and guibas @xcite gave an @xmath18 time algorithm for this problem .
et al . _
@xcite showed that this problem can be solved in @xmath19 time if a triangulation of @xmath3 is given along with @xmath3 .
since any @xmath3 can be triangulated in @xmath19 @xcite , the algorithm of guibas _ et al .
_ always runs in @xmath19 time @xcite .
another linear time solution was obtained independently by @xcite .
the @xmath20 problem in the query version has been considered by few .
it was shown in @xcite that a simple polygon @xmath3 can be preprocessed in @xmath12 time and @xmath13 space such that , given an arbitrary query line segment inside @xmath3 , @xmath21 time is required to recover @xmath22 weakly visible vertices .
this result was later improved by @xcite in which the preprocessing time and space were reduced to @xmath4 and @xmath5 respectively , at expense of more query time of @xmath23 . in a recent work , we presented an algorithm to report @xmath2 of any @xmath1 in @xmath24 time by spending @xmath12 time and @xmath13 space for preprocessing @xcite .
later , chen and wang considered the same problem and , by improving the preprocessing time of the visibility algorithm of bose _
@xcite , they improved the preprocessing time to @xmath13 @xcite . in this paper
, we show that the @xmath0 of a line segment @xmath1 can be reported in near optimal time of @xmath6 , after preprocessing the input polygon in time and space of @xmath4 and @xmath5 respectively .
compared to the algorithms in @xcite and @xcite , the storage and preprocessing time has one fewer linear factor , at expense of more query time of @xmath6 .
our approach is inspired by aronov _
algorithm for computing the visibility polygon of a point @xcite . in section [ sec : partial ] , we first show how to compute the _ partial weak visibility polygon _
@xmath25 when @xmath1 is not inside a sub - polygon @xmath26 of @xmath27 .
then , in section [ sec : balanced ] , we use a balanced triangulation to compute and report the final weak visibility polygon .
in this section , we introduce some basic terminologies used throughout the paper . for a better introduction to these terms ,
we refer the readers to guibas _ et al . _
@xcite , bose _
@xcite , and aronov _ et al .
_ @xcite . for simplicity , we assume that no three vertices of the polygon are collinear .
let @xmath3 be a simple polygon with @xmath10 vertices .
also , let @xmath28 and @xmath8 be two points inside @xmath3 .
the _ visibility sequence _ of a point @xmath28 is the sequence of vertices and edges of @xmath3 that are visible from @xmath28 .
a _ visibility decomposition _ of @xmath3 is to partition @xmath3 into a set of _ visibility regions _ , such that any point inside each region has the same visibility sequence .
this partition is induced by the _
critical constraint edges _ , which are the lines in the polygon each induced by two vertices of @xmath3 , such that the visibility sequences of the points on its two sides are different . in a simple polygon ,
the visibility sequences of two _ neighboring _ visibility regions which are separated by an edge , differ only in one vertex .
this fact is used to reduce the space complexity of maintaining the visibility sequences of the regions @xcite .
this is done by defining the _ sink regions_. a sink is a region with the smallest visibility sequence compared to all of its adjacent regions .
therefore , it is sufficient to maintain the visibility sequences of the sinks , from which the visibility sequences of all other regions can be computed . by constructing a directed dual graph over the visibility regions ( see figure [ fig : f2 ] ) , one can maintain the difference between the visibility sequences of the neighboring regions @xcite . in a simple polygon with @xmath10 vertices ,
the number of visibility and sink regions are @xmath13 and @xmath5 , respectively @xcite . here
, we present the @xmath19 time algorithm of guibas _ et al . _ for computing @xmath2 of a line segment @xmath1 inside @xmath3 , as described in @xcite
this algorithm is used in computing the partial weak visibility polygons in an output sensitive way , to be explained in section [ sec : partial : wvp ] . for simplicity , we assume that @xmath1 is a convex edge of @xmath27 , but we will show that this can be extended for any line segment in the polygon .
let @xmath29 denote the shortest path tree in @xmath3 rooted at @xmath28 .
the algorithm traverses @xmath29 using a dfs and checks the turn at each vertex @xmath30 in @xmath29 .
if the path makes a right turn at @xmath30 , then we find the descendant of @xmath30 in the tree with the largest index @xmath31 ( see figure [ fig : f1 ] ) .
as there is no vertex between @xmath32 and @xmath33 , we can compute the intersection point @xmath34 of @xmath35 and @xmath36 in @xmath37 time , where @xmath38 is the parent of @xmath30 in @xmath29 .
finally the counter - clockwise boundary of @xmath3 is removed from @xmath30 to @xmath34 by inserting the segment @xmath39 .
let @xmath26 denote the remaining portion of @xmath3 .
we follow the same procedure for @xmath8 .
this time , the algorithm checks every vertex to see whether the path makes its first left turn .
if so , we will cut the polygon at that vertex in a similar way .
after finishing the procedure , the remaining portion of @xmath26 would be the @xmath2 . .
in the left figure , the shortest path from @xmath28 to @xmath32 makes a first right turn at @xmath30 . in the right figure
, the shortest path from @xmath8 to @xmath40 makes a first left turn at @xmath41 . ]
suppose that a simple polygon @xmath3 is divided by a diagonal @xmath42 into two parts , @xmath43 and @xmath44 . for a query line segment @xmath45
, we define the _ partial weak visibility polygon _ @xmath46 ( or @xmath47 for clarity ) to be the polygon @xmath48 . in other words
, @xmath46 is the portion of @xmath3 that is weakly visible from @xmath1 _ through _ @xmath42 . in this section
, we will show how to compute @xmath46 . later in section [ sec : balanced ] , we will use this structure to compute @xmath2 .
is defined as the part of the sub - polygon @xmath43 that is weakly visible from @xmath1 . ]
we will show how to use the algorithm of guibas _ et al .
_ @xcite to compute @xmath46 . to do so
, we preprocess the polygon so that we can answer the visibility query in an output sensitive way .
the idea is to compute the visibility decomposition of the polygon and , for each decomposition cell , compute the potential shortest path tree structures .
as the number of visibility regions is @xmath13 , the preprocessing cost of our approach would be high . to overcome
, we only consider the critical constraint edges that cut @xmath42 .
the number of such constraint edges is @xmath19 and the complexity of the decomposition is reduced to @xmath5 .
this decomposition can be computed in @xmath5 time .
we call this decomposition the _ partial visibility decomposition _ of @xmath3 with respect to @xmath42
. the remaining part of this section shows how to modify the linear algorithm of guibas _ et al . _
@xcite so that @xmath46 can be computed in an output sensitive way .
first , we show how to compute the shortest path trees , and then present our algorithm for computing @xmath46 .
we define the _ partial shortest path tree _
@xmath49 to be the subset of @xmath29 that lead to a leaf node in @xmath43 . in other words
, @xmath49 is the union of the shortest paths from @xmath28 to all the vertices of @xmath43 . in this section ,
we show how to preprocess the polygon @xmath27 , so that for any given point @xmath50 , any part of @xmath49 can be traversed in an output sensitive way . the shortest path tree @xmath49 is composed of two kinds of edges : the _ primary edges _ that connect the root @xmath28 to its direct visible vertices , and the _ secondary edges _ that connect two vertices of @xmath49 ( see figure [ fig : pspt ] ) .
we also recognize two kinds of secondary edges : the 1st type of secondary edges ( 1st type for short ) are those edges that are connected to a primary edge , and the 2nd type are the ones that connect other vertices of the polygon .
notice that if a point @xmath28 crosses a critical constraint and that constraint does not cut @xmath42 , then the structure of @xmath51 would not change . for different points of @xmath44 .
notice that as @xmath8 and @xmath52 are on the same visibility region w.r.t .
@xmath43 , @xmath53 and @xmath54 have the same structure . ]
we can compute the primary edges of @xmath55 by using aronov s output sensitive algorithm of computing the partial visibility polygon @xcite .
more precisely , with a processing cost of @xmath4 time and @xmath5 space , giving a point @xmath28 in query time , a pointer to the sorted list of the vertices that are visible to @xmath28 can be computed in @xmath56 time .
it is also necessary to compute the list of the secondary edges of every vertex of @xmath57 .
each vertex @xmath52 in @xmath57 have @xmath19 possible 2nd type edges .
depending on the parent of @xmath52 , a sub - list of these edges would appear in @xmath55 . to store all the possible 2nd type edges of @xmath52 ,
we compute and store this sub - list , or to be precise , the starting and ending edges of the list , for all the possible parents of @xmath52 .
as there are @xmath19 possible parents for a vertex , these calculations can be performed for all the vertices of the polygon in total time of @xmath4 and the data can be stored in @xmath5 space .
having these data , we can , in the query time , access the list of the 2nd type edges of any vertex in constant time .
we build the same structure for the 1st type edges .
the parent of a 1st type edge is the root of the tree . as the root can be in any of the @xmath5 different visibility regions , computing and storing the starting and ending edges in the list of 1st type edges of a vertex cost @xmath12 time and @xmath13 space .
we can reduce the time and space needed to compute and store these structures , having this property that two adjacent regions have only @xmath37 differences in their 1st type edges .
[ lemm : lemma2 ] consider a visibility region @xmath58 in the polygon and suppose that the 1st type secondary edges are computed for a point @xmath28 in this region .
for a neighboring region that share a common edge with @xmath58 , these edges can be updated in constant time .
when a view point @xmath28 crosses the border of two neighboring regions , a vertex becomes visible or invisible to @xmath28 @xcite . in figure
[ fig : h4 ] for example , when @xmath28 crosses the border specified by @xmath59 and @xmath15 , a 1st type secondary edge of @xmath59 becomes a primary edge of @xmath28 , and all the edges of @xmath15 become the 1st type secondary edges .
we can see that no other vertex is affected by this movement .
processing these changes can be done in constant time , since it includes the following changes : removing a secondary edge of @xmath59 ( @xmath60 ) , adding a primary edge ( @xmath61 ) , and moving an array pointer ( edges of @xmath15 ) from the 2nd type edges of @xmath60 to the 1st type edges of @xmath61 .
note that we know the exact positions of these elements in their corresponding lists .
finally , the only edge which involves in these changes can be identified in the preprocessing time ( the edge corresponding to the crossed critical constraint ) , so , the time we spent in the query time would be @xmath37 .
enters a new visibility region , the combinatorial structure of @xmath49 can be maintained in constant time . ]
having this fact and using a _ persistent data structure _ , e.g. _ persistent red - black tree _
@xcite , we can reduce the cost of storing the 1st type edges by a linear factor .
a persistent red - black tree is a red - black tree that can remember all its intermediate versions .
if a set of @xmath10 linearly ordered items are stored it the tree and we perform @xmath62 update into it , any version @xmath63 , for @xmath64 , can be retrieved in time @xmath56 .
this structure can be constructed in @xmath65 time by using @xmath66 space .
[ theo : theo - pspt ] a simple polygon @xmath27 can be processed into a data structure with @xmath5 space and in @xmath4 time so that for any query point @xmath28 , the shortest path tree from @xmath28 can be reported in @xmath67 , where @xmath22 is the size of the tree that is to be reported .
first , we use aronov s algorithm for computing the partial visibility polygon of @xmath28 . for this , @xmath5 space and @xmath4 time is needed in the preprocessing phase . for the secondary edges , @xmath4 time and @xmath5 space
is needed to compute and store these edges .
also , a point location structure is built on top of the arrangement . in the query time
, the partial visibility region of @xmath28 can be located in @xmath56 to have the sorted list of the visible vertices from @xmath28 .
as the visible vertices from @xmath28 correspond to the primary edges of @xmath55 , we also have the primary edges of @xmath49 . for the 1st type edges ,
a tour is formed to visit all the cells of the partial visibility decomposition . from lemma [ lemm :
lemma2 ] , we can start from an arbitrary cell , walk along the tour , and construct a persistent red - black tree on the 1st type edges of @xmath55 of a point in each cell .
as there are @xmath5 cells and , each cell has @xmath19 1st type edges , the structure takes @xmath5 storage and can be built in @xmath4 preprocessing time . having this structure ,
the 1st type edges of the cell containing @xmath28 can be retrieved from the persistent data structure in @xmath56 time .
finally , at each node of the tree , we have the list of 2nd type edges from that node .
therefore , the cost of traversing @xmath55 is the number of visited nodes of the tree , plus the initial @xmath56 time .
in other words , the query time is @xmath67 , where @xmath22 is the number of the traversed edges of the @xmath55 .
now that that we show how to compute @xmath51 for any point @xmath50 in the query time , we can use the linear algorithm presented in section [ sec : guibas ] for computing @xmath0 of a simple polygon and modify it to compute @xmath46 in an output sensitive way .
as we can see in figure [ fig : polygonal - edge ] , the algorithm can be extended to the cases that @xmath1 is not a polygonal edge .
we can assume @xmath1 to be a polygonal edge . ] to achieve an output sensitive algorithm , we store some additional information about the vertices of the polygon in the preprocessing phase .
we say that a vertex @xmath68 is _ left critical _ ( lc for short ) with respect to a point @xmath69 , if @xmath70 makes its first left turn at @xmath15 or one of its ancestors . in other words , each shortest path from @xmath28 to a non - lc vertex
is a convex chain that makes only clockwise turns at each node .
the _ critical state _ of a vertex is whether it is lc or not .
if we have the critical state of all the vertices of @xmath43 with respect to a point @xmath8 , we say that we have the _ critical information _ of @xmath8 .
the idea is to change the algorithm of section [ sec : guibas ] and make it output sensitive .
the outline of the algorithm is as follows : in the first round , we traverse @xmath51 using dfs . at each vertex
, we check whether this vertex is left critical with respect to @xmath8 or not .
if so , we are sure that the descendants of this vertex are not visible from @xmath1 , so , we postpone its processing to the time it is reached from @xmath8 , and check other branches of @xmath51 .
otherwise , proceed with the algorithm and check whether @xmath51 makes a right turn at this vertex . in the second round ,
we traverse @xmath71 and perform the normal procedure of the algorithm .
[ lemm : lemma3 ] all the traversed vertices in @xmath51 and @xmath71 are vertices of @xmath46 . in the preprocessing phase ,
we compute the critical information of a point inside each region , and assign this information to that region . in the query time and upon receiving a line segment @xmath1 , we find the regions of @xmath28 and @xmath8 . using the critical information of these two regions , the above algorithm can be applied to compute @xmath46 . as there are @xmath5 visibility regions in the partial visibility decomposition , @xmath13 space is needed to store the critical information of all the vertices . for each region
, we compute @xmath55 of a point , and by traversing the tree , we update the critical information of each vertex with respect to this region .
an array of size @xmath19 is assigned to each region to store these information .
we also build the structure described in section [ sec : pspt ] to compute @xmath72 in @xmath12 time and @xmath13 space . in the query time
, we locate the visibility regions of @xmath28 and @xmath8 in @xmath56 time . by remark [ lemm : lemma3 ] ,
when we proceed the algorithm in @xmath55s of @xmath28 and @xmath8 , we only traverse the vertices of @xmath46 .
finally , as the processing time spent in each vertex is @xmath37 , the total query time is @xmath73 . to improve this result
, we use the fact that any two adjacent regions have @xmath37 differences in their critical information .
[ lemm : lemma5 ] in the path between neighboring visibility regions , the changes of the critical information can be handled in constant time .
0.4 w.r.t .
@xmath28 , as @xmath28 moves between the two regions.,title="fig : " ] 0.4 w.r.t .
@xmath28 , as @xmath28 moves between the two regions.,title="fig : " ] + 0.4 w.r.t .
@xmath28 , as @xmath28 moves between the two regions.,title="fig : " ] 0.4 w.r.t .
@xmath28 , as @xmath28 moves between the two regions.,title="fig : " ] suppose that we want to maintain the critical information of @xmath28 and @xmath28 is crossing the critical constraint defined by @xmath60 , where @xmath59 and @xmath15 are two reflex vertices of @xmath27 .
the only vertices that affect directly by this change are @xmath59 and @xmath15 .
depending on the critical states of @xmath59 and @xmath15 w.r.t .
@xmath28 , four situations may occur ( see figure [ fig : e2 - 1 ] ) . in the first three cases ,
the critical state of @xmath15 will not change . in the forth case ,
however , the critical state of @xmath15 will change . before the cross
, the shortest path makes a left turn at @xmath59 , therefore , both @xmath59 and @xmath15 are lc w.r.t .
however , after the cross , @xmath15 is no longer lc .
this means that the critical state of all the children of @xmath15 in the @xmath49 may change as well . to handle these cases
, we modify the way the critical information of each vertex w.r.t .
@xmath28 are stored . at each vertex @xmath15 , we store two additional values : the number of lc vertices we met in the path @xmath74 from @xmath28 , or its _ critical number _ , and a _ debit number _ , which is the critical number that is to be propagated in the vertex subtree .
it is clear that if a vertex is lc , it means that its critical number is greater than zero ( see figure [ fig : crtitical - numbers ] ) .
also , if a vertex has a debit number , the critical numbers of all its children must be added by this debit number .
notice that computing and storing these numbers along the critical information will not change the time and space requirements . in @xmath75 .
] now let us consider the forth case in figure [ fig : e2 - 1 ] .
when @xmath15 becomes visible to @xmath28 , it is no longer lc w.r.t .
therefore , the critical number of @xmath15 is changed to @xmath76
. however , instead of changing the critical numbers of all the children of @xmath15 , we set the debit number of @xmath15 to -1 , indicating that the critical numbers of all the vertices of its subtree must be subtracted by 1 .
the actual propagation of this subtraction will happen at query time when @xmath49 is traversed .
if @xmath28 moves in the reverse path , i.e. , when @xmath15 becomes invisible to @xmath28 , we handle the tree in the same way by storing @xmath77 in the debit numbers , and propagating this addition in the query time .
a persistent data structure can be used to reduce the costs to @xmath4 preprocessing time and @xmath5 storage .
we form a tour visiting all the cells and construct a persistent red - black tree on the critical information and the 2nd type edges of all the nodes .
the structure takes @xmath5 storage and can be built in @xmath4 preprocessing time .
in addition , we build a point location structure on top of the arrangement which can be done in @xmath5 time and @xmath5 space @xcite . [
theorem : partial - visibility ] given a polygon @xmath27 and a diagonal @xmath42 which cuts @xmath27 into two parts , @xmath43 and @xmath44 , and using @xmath4 time , we can construct a data structure of size @xmath5 so that , for any query line segment @xmath45 , the partial weak visibility polygon @xmath46 can be reported in @xmath73 time .
there is always a diagonal @xmath42 of a simple polygon that cuts @xmath3 into two pieces , each having at most @xmath78 vertices @xcite .
we can recursively subdivide and build a balanced binary tree where the leaves are triangles and each interior node @xmath79 corresponds to a subpolygon @xmath80 and a diagonal @xmath81 .
each diagonal @xmath81 divides @xmath80 into two subpolygons , @xmath82 and @xmath83 , which respectively correspond to the left and right subtrees of @xmath79 ( see figure [ fig : triangulation ] ) .
we build the data structures described in section [ sec : partial ] for @xmath82 and @xmath83 with respect to @xmath81 .
[ cross/.style = draw = red , thick , circle , fill = red!20,minimum height=1em , leaf/.style = draw = blue , thick , circle , fill = blue!20,minimum height=1em , normal/.style = draw = yellow!4 , thick , circle , fill = yellow!10,minimum height=1em , level distance=8 mm , every node/.style = circle , inner sep=1pt , level 1/.style = sibling distance=32 mm , level 2/.style = sibling distance=16 mm , level 3/.style = sibling distance=8 mm , level 4/.style = sibling distance=4 mm , level 5/.style = sibling distance=2 mm ] child node [ cross ] 2 child node 3 child node 4 child node 5 child node 5 child node 4 child node 5 child node 5 child node [ cross ] 3 child node 4 child node 5 child node 5 child node [ cross ] 4 child node 5 child node [ leaf ] 5 child node [ cross ] 2 child node [ cross ] 3 child node [ cross]4 child node [ leaf ] 5 child node [ leaf ] 5 child node [ leaf]4 child node 5 child node 5 child node [ cross ] 3 child node [ cross]4 child node [ leaf]5 child node 5 child node 4 child node 5 child node 5 ; to compute @xmath2 , @xmath28 and @xmath8 will be located among the leaf triangles . in the simplest case ,
both @xmath28 and @xmath8 belong to the same triangle .
first we explain this situation .
we construct @xmath84 for each @xmath79 from the leaf to the root . here , @xmath84 is the partial weak visibility polygon of @xmath1 in @xmath80 with respect to @xmath81 . for the leaf node ,
it is the corresponding triangle , and for other nodes , it can be computed inductively . in this case
, the inductive step is similar to that of @xcite . in each step
, the merging of the computed polygons can be done in @xmath56 .
the time and space needed for building an exterior visibility decomposition of a simple polygon with @xmath62 vertices are @xmath85 and @xmath86 , respectively .
thus , the space and time of the above inductive procedure can be expressed as the following equations : @xmath87 therefore , @xmath88 , and @xmath89 . with the same analysis as in @xcite
, we can calculate the query time .
two point locations can be done in @xmath56 time .
as the triangulation is balanced , we path from the root to any node has @xmath56 length . as we showed in theorem [ theorem : partial - visibility ] , the time needed to query @xmath84 at step @xmath79 is @xmath90 .
also , the merging at each step can be done in @xmath56 time .
therefore , the total query time is @xmath91 , or @xmath92 .
the tricky part is when @xmath28 and @xmath8 are on different triangles .
assume that at step @xmath79 , the query line segment is @xmath93 and it is in sub - polygon @xmath80 .
the sub - polygon @xmath80 is divided by diagonal @xmath81 to two sub - polygons @xmath82 and @xmath83 .
if @xmath93 does not intersect @xmath81 , without loss of generality , assume that @xmath93 is located in @xmath83 ( see figure [ fig : inductive - step-1 ] ) .
we do the normal procedure of the algorithm and compute @xmath94 .
we continue to recursively compute weak visibility polygon on @xmath83 . in this case
, the time needed by this step can be expressed as @xmath95 . on the other hand , if @xmath93 and @xmath81 intersect at point @xmath96 , without loss of generality , assume that @xmath97 is in @xmath83 and @xmath98 is in @xmath82 ( see figure [ fig : inductive - step-2 ] ) .
we can express @xmath99 as the union of four weak visibility polygons : 1 .
the partial weak visibility polygon of @xmath100 on @xmath82 , 2 .
the weak visibility polygon of @xmath100 in @xmath83 , 3 .
the partial weak visibility polygon of @xmath101 on @xmath83 , 4 .
the weak visibility polygon of @xmath101 in @xmath82 . in other words ,
we must compute two partial weak visibility polygons , and two weak visibility sub - problems . having these four visibility polygons ,
the union of them can be merged in time @xmath102 . according to the theorem [ theorem : partial - visibility ] , the query time spent at step @xmath79
can be expressed as : @xmath103 .
0.48 0.48 the preprocessing costs of the algorithm is the same as before . for the query time , by induction , we can prove that @xmath104 .
assume that this property holds for every simple polygon with less than @xmath10 vertices and every line segment @xmath1 in it .
for a simple polygon @xmath27 with @xmath10 vertices , depending on whether @xmath1 cuts the diagonal @xmath42 , we have one of these equation : @xmath105 here , @xmath106 is the partial weak visibility polygon of @xmath1 w.r.t . the cut @xmath42 . in the first case
, we have @xmath107 in the second case , we have @xmath108 in these equations , @xmath109 is the weak visibility polygon of @xmath1 in a polygon of size @xmath110 . as for the base case
, we showed that if @xmath1 is located in single triangle , the time for computing @xmath2 would be @xmath6 . in summary
, we have the following theorem : [ theoremxx ] a simple polygon @xmath3 of size @xmath10 can be processed in @xmath4 time into a data structure of size @xmath5 so that , for any query line segment @xmath1 , @xmath2 can be reported in time @xmath111 .
in this paper , we showed how to answer the weak visibility queries in a simple polygon with @xmath10 vertices in an efficient way . in the first part of the paper
, we defined the partial weak visibility polygon @xmath112 of a line segment @xmath1 with respect to a diagonal @xmath42 and presented an algorithm to report it in time @xmath113 , by spending @xmath4 time to preprocess the polygon and maintaining a data structure of size @xmath5 .
in the second part , we presented a data structure of size @xmath5 which can be computed in time @xmath4 so that the weak visibility polygon @xmath2 from any query line segment @xmath114 can be reported in time @xmath115 .
b. aronov , l. j. guibas , m. teichmann and l. zhang .
visibility queries and maintenance in simple polygons . _ discrete and computational geometry _ , 27(4):461 - 483 , 2002 .
p. bose , a. lubiw , j. i. munro .
efficient visibility queries in simple polygons .
_ computational geometry : theory and applications _ , 23(3):313 - 335 , 2002 | this paper considers the problem of computing the weak visibility polygon ( @xmath0 ) of any query line segment @xmath1 ( or @xmath2 ) inside a given simple polygon @xmath3 .
we present an algorithm that preprocesses @xmath3 and creates a data structure from which @xmath2 is efficiently reported in an output sensitive manner .
our algorithm needs @xmath4 time and @xmath5 space in the preprocessing phase to report @xmath2 of any query line segment @xmath1 in time @xmath6 .
we improve the preprocessing time and space of current results for this problem @xcite at the expense of more query time . computational geometry , visibility , line segment visibility |
let @xmath6 be a compact set in the euclidean space @xmath7 the _ covariogram _ @xmath8 of @xmath6 is the function on @xmath2 defined by @xmath9 where @xmath10 stands for the @xmath11-dimensional lebesgue measure .
this function , which was introduced by g. matheron in his book ( * ? ? ?
* section 4.3 ) on random sets , is also called _ set covariance_.
the covariogram @xmath8 coincides with the _ autocorrelation _ of the characteristic function @xmath12 of @xmath13 i.e. : @xmath14 the covariogram @xmath8 is clearly unchanged with respect to translations and reflections of @xmath6 , where , throughout the paper , _ reflection _ means reflection in a point .
a _ convex body _ in @xmath2 is a convex compact set with nonempty interior . in 1986
matheron @xcite asked the following question and conjectured a positive answer for the case @xmath5 .
does the covariogram determine a convex body in @xmath15 among all convex bodies , up to translations and reflections ?
we are able to confirm matheron s conjecture completely .
[ matconjconf ] every planar convex body is determined within all planar convex bodies by its covariogram , up to translations and reflections . it is known that the covariogram problem is equivalent to any of the following problems . 1 . [ chord_lengths ]
determine a convex body @xmath1 by the knowledge , for each unit vector @xmath16 in @xmath2 , of the distribution of the lengths of the chords of @xmath1 parallel to @xmath16 .
[ distribution_xy ] determine a convex body @xmath1 by the distribution of @xmath17 , where @xmath18 and @xmath19 are independent random variables uniformly distributed over @xmath1 .
[ phase_retrieval ] determine the characteristic function @xmath20 of a convex body @xmath1 from the modulus of its fourier transform @xmath21 .
in view of theorem [ matconjconf ] , for problems p[chord_lengths ] and p[distribution_xy ] the determination holds within the class of planar convex bodies and for problem p[phase_retrieval ] within the class of characteristic functions of planar convex bodies . in each of the three problems the determination is unique up to translations and reflections of the body .
the equivalence of the covariogram problem and p[chord_lengths ] was observed by matheron , who showed in @xcite that the derivatives @xmath22 for all @xmath23 , yield the distribution of the lengths of the chords of @xmath1 parallel to @xmath16 .
blaschke @xcite asked whether the distribution of the lengths of _ all _ chords ( that is , not separated direction by direction ) of a planar convex body determines that body , up to isometries in @xmath24 mallows and clark @xcite constructed polygonal examples that show that the answer is negative in general .
gardner , gronchi , and zong @xcite observed that the distribution of the lengths of the chords of @xmath1 parallel to @xmath16 coincides , up to a multiplicative factor , with the _ rearrangement _ of the x - ray of @xmath1 in direction @xmath16 , and rephrased p[chord_lengths ] in these terms .
chord - length distributions are of wide interest beyond mathematics , as mazzolo , roesslinger , and gille @xcite describe .
see also schneider @xcite and cabo and baddeley @xcite .
problem p[distribution_xy ] was asked by adler and pyke @xcite in 1991 ; see also @xcite .
its equivalence to the covariogram problem comes from the observation that the convolution in is the probability density of @xmath17 , up to a multiplicative factor .
problem p[phase_retrieval ] is a special case of the _ phase retrieval problem _
, where @xmath20 is replaced by a function with compact support .
the phase retrieval problem has applications in _ x - ray crystallography , optics , electron microscopy _ and other areas , references to which may be found in @xcite .
the equivalence of the covariogram problem and p[phase_retrieval ] follows by applying the fourier transform to and using the relation @xmath25 .
recently , baake and grimm @xcite have observed that the covariogram problem is relevant for the inverse problem of finding the atomic structure of a _ quasicrystal _ from its _ x - ray diffraction image_.
it turns out that quasicrystals can often be described by means of the so - called _ cut - and - project scheme _ ; see @xcite . in this scheme a quasiperiodic discrete subset @xmath26 of @xmath15 which models the atomic structure of a quasicrystal ,
is described as the canonical projection of @xmath27 onto @xmath15 where @xmath28 ( which is called _ window _ ) is a subset of @xmath29 and @xmath30 is a lattice in @xmath31 for many quasicrystals , the lattice @xmath30 can be recovered from the diffraction image of @xmath32 thus , in order to determine @xmath33 it is necessary to know @xmath28 .
the covariogram problem enters at this point , since @xmath34 can be obtained from the diffraction image of @xmath32 note that the set @xmath28 is in many cases a convex body . in ( * ? ? ?
* theorem 6.2 and question 6.3 ) the covariogram problem was transformed to a question for the so - called _ radial mean bodies_. a planar convex body @xmath1 can be determined by its covariogram in a class @xmath35 of sets which is much larger than that of convex bodies .
this is a consequence of theorem [ matconjconf ] and of a result of benassi , bianchi , and dercole @xcite . in @xcite
the class @xmath35 is defined and it is proved that a body @xmath36 whose covariogram is equal to that of a convex body is necessarily convex . however , in theorem [ matconjconf ] the assumption that @xmath1 is convex is crucial , since there exist examples of non - convex sets which are neither translations nor reflections of each other and have equal covariograms ; see @xcite , rataj @xcite , and @xcite . the first partial solution of matheron s conjecture was given by nagel @xcite in 1993 , who confirmed it for all convex polygons .
schmitt @xcite , in the same year , gave a constructive proof of the determination of each set in a suitable class of polygons by its covariogram .
this class contains each convex polygon without parallel edges and also some non - convex polygons . in 2002
bianchi , segala and voli @xcite gave a positive answer to the covariogram problem for all planar convex bodies whose boundary has strictly positive continuous curvature .
bianchi @xcite proved a common generalization of this and nagel s result .
in @xcite the authors of this paper studied how much of the covariogram data is needed for the uniqueness of the determination , and also extended the class of bodies for which the conjecture was confirmed .
the covariogram problem in the general setting has negative answer , as bianchi @xcite proved by finding counterexamples in @xmath2 for every @xmath37 . for other results in dimensions higher that two we refer to goodey , schneider , and weil @xcite , and @xcite . in @xcite
it is proved that a convex three - dimensional polytope is determined by its covariogram .
this proof requires the following generalization of the covariogram problem .
cross covariogram _ of two convex bodies @xmath1 and @xmath38 in @xmath39 is the function defined for each @xmath40 by @xmath41 bianchi @xcite proves that if @xmath1 and @xmath38 are convex polygons , then @xmath42 determines _ both @xmath1 and @xmath38 _ , with exclusion of a completely described family of exceptions .
the family of exceptions is composed of pairs of parallelograms .
in view of results from @xcite , for proving theorem [ matconjconf ] it suffices to derive the following statement .
[ arcdetermprop ] let @xmath1 and @xmath38 be planar strictly convex and @xmath43 regular bodies with equal covariograms .
then @xmath38 possesses a non - degenerate boundary arc whose translation or reflection lies in the boundary of @xmath44 theorem [ matconjconf ] follows directly from proposition [ arcdetermprop ] and the following two statements . _
( bianchi @xcite ) _ [ irregthm ] let @xmath1 and @xmath38 be planar convex bodies with equal covariograms .
assume that one of them is not strictly convex or not @xmath43 regular .
then @xmath1 and @xmath38 are translations or reflections of each other . _
( bianchi @xcite ) _ [ fromarcstobodies ] let @xmath1 and @xmath38 be planar convex bodies with equal covariograms and a common non - degenerate boundary arc .
then @xmath1 and @xmath38 coincide , up to translations and reflections .
the `` heart '' of the proof is contained in section [ sect : rel : pos : par ] , which also contains , in the beginning , an explanation of the main ideas .
many natural questions for the covariogram problem are still open .
we mention here some of them . 1 . which four - dimensional convex polytopes are determined by their covariogram ? 2
all known examples of convex bodies that are not determined by their covariogram are cartesian products .
do there exist other examples ?
3 . is the answer to the covariogram problem positive for all three - dimensional convex bodies whose boundary has continuous and strictly positive principal curvatures ?
the closure , boundary , interior , linear hull , affine hull , and convex hull of a set , and the support of a function , are abbreviated , in the standard way , by @xmath45 @xmath46 @xmath47 @xmath48 @xmath49 @xmath50 and @xmath51 respectively .
we denote by @xmath52 @xmath53 @xmath54 and @xmath55 origin , scalar product , euclidean norm , and euclidean unit sphere in @xmath15 respectively . in analytic expressions elements of @xmath2
are identified with real column vectors of length @xmath56 thus , @xmath57 where @xmath58 denotes matrix transposition . throughout the paper we use the matrix @xmath59 of 90-degree rotation in the counterclockwise orientation .
we do not distinguish between @xmath60 matrices over @xmath61 and linear operators in @xmath24 for vectors @xmath62 we put @xmath63 where @xmath64 stands for the matrix whose first column is @xmath65 and the second one is @xmath66 regarding standard notations and notions from the theory of convex sets we mostly follow the monograph @xcite .
the _ difference body _ of a convex body @xmath1 is the set @xmath67 it is not hard to see that @xmath68 a boundary point @xmath69 of a convex body @xmath1 is said to be @xmath43 _ regular _ if there exists precisely one hyperplane supporting @xmath1 at @xmath70 furthermore , a convex body @xmath1 is said to be @xmath43 _ regular _ if all boundary points of @xmath1 are @xmath43 regular .
we say that a convex polygon @xmath71 is _ inscribed _ in a convex body @xmath72 if all vertices of @xmath73 lie in @xmath74 given @xmath75 , @xmath76 , the chord @xmath77 $ ] is said to be an _ affine diameter _ of @xmath1 , if for some @xmath78 the vectors @xmath16 and @xmath79 are outward normals of @xmath1 at @xmath75 and @xmath80 respectively .
it is well known that @xmath77 $ ] is an affine diameter of @xmath1 if and only if@xmath81 if @xmath1 is a planar convex body and @xmath82 are two distinct boundary points of @xmath83 then @xmath84_{k}$ ] stands for the counterclockwise boundary arc of @xmath1 starting at @xmath69 and terminating at @xmath85
let @xmath1 be a strictly convex and @xmath43 regular convex body in @xmath24 consider an arbitrary @xmath86 then there exist points @xmath87 in counterclockwise order on @xmath83 such that @xmath88 see fig .
[ p : def : fig:1 ] , also regarding notations introduced below .
then the set @xmath89 is a parallelogram inscribed in @xmath83 whose edges are translates of @xmath90 $ ] and @xmath91 $ ] with @xmath92 by @xmath93 we denote the outward unit normal of @xmath1 at @xmath94 cc1 mm ( 54,50 ) ( -1,-0.7 ) ( 51,29)@xmath1 ( 36,14)@xmath95 ( 34,35)@xmath96 ( 9,9)@xmath97 ( 7,30)@xmath98 ( 46,5)@xmath99 ( 43,45)@xmath100 ( -1,40)@xmath101 ( 7,-3)@xmath102 ( 22,10.5)@xmath65 ( 31.7,24)@xmath103 + for the sake of brevity , dealing with functionals @xmath104 depending on @xmath1 and @xmath105 we shall also use the notations @xmath106 or @xmath107 instead of @xmath108 provided the choice of @xmath1 and/or @xmath65 is clear from the context .
[ frst : drv : cov : thm ] let @xmath1 be a strictly convex and @xmath43 regular body in @xmath39 and let @xmath86 then the following statements hold .
a. [ dreprstat ] the covariogram @xmath0 is continuously differentiable at @xmath109 moreover , @xmath110 b. [ pdcontcond ] the functions @xmath111 with @xmath112 are continuous at @xmath109 c. [ ptrans : prop ] for every strictly convex @xmath43 regular body @xmath38 with @xmath113 the parallelogram @xmath114 is a translate of @xmath115 part [ dreprstat ] is known ; see and .
let us prove part [ pdcontcond ] . if @xmath116 then @xmath117 and @xmath118 intersect precisely at @xmath119 and @xmath120 furthermore , the intersection is transversal . by the implicit function theorem ,
this implies that @xmath119 and @xmath121 depend continuously on @xmath109 since @xmath1 is @xmath43 regular , the outward unit normal @xmath122 of @xmath1 at a boundary point @xmath69 of @xmath1 depends continuously on @xmath70 therefore , for @xmath123 the function @xmath124 depends continuously on @xmath109 part [ ptrans : prop ] follows directly from . in what follows ,
notations involving an integer subscript @xmath125 ranging in a certain interval are extended periodically to all @xmath126 for example , we set @xmath127 where @xmath128 @xmath129 and @xmath130 throughout the paper , the parallelograms @xmath131 will provide a convenient geometric representation of some information contained in the covariogram , in view of . a priori , for two planar convex bodies @xmath1 and @xmath38 with @xmath132 , the translation that carries @xmath131 to @xmath114 may depend on @xmath109 one crucial step of the proof is to show that the above translation is in fact independent of @xmath109 see the beginning of section [ sect : rel : pos : par ] for a brief sketch of the mentioned argument .
if the covariogram of @xmath1 is twice differentiable at @xmath105 we introduce the hessian matrix @xmath133_{i , j=1}^2.\ ] ] the relations given in theorem [ scnd : drv : cov : thm ] are reformulations of the relations presented in @xcite . part [ grel ] of theorem [ scnd : drv : cov : thm ] is extended to every dimension in @xcite .
we omit the proof of part [ grel ] and present a short proof of parts [ deltarel ] and [ ueq ] .
[ scnd : drv : cov : thm ] let @xmath1 be a strictly convex and @xmath43 regular body in @xmath24 then @xmath134 is continuously differentiable at every @xmath86 furthermore , for every @xmath116 the following statements hold true .
a. [ grel ] the hessian @xmath135 can be represented by @xmath136 b. [ deltarel ] the determinant of @xmath135 depends continuously on @xmath65 and satisfies @xmath137 c. [ ueq ] the vectors @xmath100 , @xmath102 and the matrix @xmath138 are related by @xmath139 _ part [ grel ] : _ for the proof see @xcite and @xcite . _
part [ deltarel ] : _ from we get @xmath140 the above two equalities imply @xmath141 taking determinants of the left and the right hand side we obtain @xmath142 let us notice that @xmath143 for every @xmath144 for instance , @xmath145 is positive because , if @xmath146 denotes the unit outward normal to the edge @xmath147 $ ] of @xmath148 then @xmath149 and @xmath150 are in this counterclockwise order on @xmath151 by the strict convexity of @xmath44 by , we may divide by @xmath152 , arriving at the equality in .
the inequality in follows from .
equality follows directly from and the algebraic identity @xmath153 which holds for all @xmath154 and can be found , in a much more general form , in .
the continuity of @xmath155 is a consequence of and theorem [ frst : drv : cov : thm ] . _
part [ ueq ] : _ we multiply by @xmath156 from the left and by @xmath157 from the right getting @xmath158 expressing the entries of @xmath159 over the entries of @xmath160 one can see that @xmath161 hence , taking into account that @xmath162 we arrive at .
[ central symmetry ] let @xmath1 be a strictly convex and @xmath43 regular body in @xmath39 .
the following conditions are equivalent .
a. [ cs_one ] the body @xmath1 is centrally symmetric .
b. [ cs_two ] at least one diagonal of each parallelogram inscribed in @xmath1 is an affine diameter of @xmath1 . c. [ cs_three ]
the covariogram @xmath0 is a solution of the _ monge - ampre differential equation _
@xmath163 for @xmath164 .
the implication @xmath165 is trivial .
the equivalence of and follows from .
it remains to prove that implies .
let us first prove that implies that both diagonals of each parallelogram inscribed in @xmath1 are affine diameters . assume the contrary .
then , for some @xmath166 , exactly one diagonal of @xmath167 , say @xmath168 $ ] , is an affine diameter .
let @xmath169 , @xmath170 $ ] , be a continuous parametrization of a small arc of @xmath117 with @xmath171 . if we define @xmath172 then @xmath173 and @xmath174 .
we claim that there exists a sufficiently small @xmath175 such that no diagonal of @xmath176 is an affine diameter of @xmath1 .
in fact , @xmath177 $ ] is not an affine diameter , because it is close to @xmath178,$ ] which is not an affine diameter . on the other hand , assume that there exists @xmath179 such that for each @xmath180 $ ] the diagonal @xmath181 $ ] is an affine diameter . since @xmath119 is the only point of @xmath117 with unit outer normal opposite to the one in @xmath182 and @xmath183 , we have @xmath181=[p_1(x),p_3(x)]$ ] .
thus , the reflection of @xmath184 \right\}}$ ] about the midpoint of @xmath168 $ ] is contained in @xmath117 , and this implies that @xmath178 $ ] is an affine diameter too , a contradiction .
this proves the claim .
it remains to prove that if both diagonals of each parallelogram inscribed in @xmath1 are affine diameters , then @xmath1 is centrally symmetric .
let @xmath166 , let @xmath169 , @xmath170 $ ] , be a continuous parametrization of the arc @xmath168_k$ ] and let @xmath185 be as above .
arguing as above one can prove that , for each @xmath186 , we have @xmath181=[p_1(x),p_3(x)]$ ] .
therefore , for each @xmath186 , @xmath169 and its reflection about the midpoint @xmath187 of @xmath168 $ ] belong to @xmath117 .
thus @xmath1 is centrally symmetric with respect to @xmath187 .
the crucial point of the proof of proposition [ arcdetermprop ] is the statement that outer normals of @xmath1 are determined by @xmath0 , up to the ambiguities arising from reflections of the body .
more precisely , we need to prove the following .
let us sketch the proof of proposition [ u1u3:determ ] .
first , we prove that there is @xmath192 such that @xmath131 and @xmath193 have the opposite vertices @xmath96 and @xmath97 in common and @xmath194 $ ] is not an affine diameter .
this clearly implies that @xmath195 and @xmath196 thus , @xmath100 and @xmath102 satisfy the system given by the two equations obtained by evaluating at both @xmath65 and @xmath66 using the geometric interpretation of the action of @xmath160 contained in lemma [ hessquadinterp ] , in lemma [ u1u3eigv ] we express the vectors @xmath100 and @xmath102 in terms of the eigenvectors of @xmath197 in order to make this expression of @xmath100 and @xmath102 dependent only on the covariogram , it remains to prove that the property that @xmath131 and @xmath193 share a diagonal is preserved across bodies with equal covariograms .
the latter is done in proposition [ hexdetermprop ] .
let us now sketch how proposition [ arcdetermprop ] follows from proposition [ u1u3:determ ] .
let @xmath1 and @xmath38 be strictly convex @xmath43 regular bodies with @xmath132 , and let us choose @xmath198 such that @xmath199 .
we will prove the following claim : _ if @xmath65 belongs to a suitable neighborhood @xmath200 of @xmath201 , and if @xmath131 and @xmath202 share their vertex @xmath97 ( i.e. @xmath203 ) , then also @xmath114 and @xmath204 share their vertex @xmath97 .
_ indeed , proposition [ u1u3:determ ] together with some continuity argument allows us to prove that when @xmath65 is close to @xmath201 and @xmath205 then we have @xmath206 . in view of the strict convexity of @xmath1 and @xmath38
, this implies the claim above .
let now @xmath185 , for @xmath170 $ ] , be a parametrization of a curve contained in @xmath200 with the property that , for each @xmath207 $ ] the parallelograms @xmath202 and @xmath208 share their vertex @xmath97 .
the previous claim implies that the arc of @xmath117 spanned by the vertex @xmath209 when @xmath186 varies in @xmath210 $ ] , is a translate of the arc of @xmath211 spanned by the vertex @xmath212 .
therefore , up to translations , @xmath117 and @xmath211 have an arc in common . [ hessquadinterp ]
let @xmath1 be a strictly convex and @xmath43 regular body in @xmath39 and let @xmath86 let @xmath213 be such that the vectors @xmath214 and @xmath100 are in counterclockwise order on @xmath215 consider the convex quadrilateral @xmath216 with consecutive vertices @xmath217 such that @xmath218 , @xmath219 and , for each @xmath220 the vector @xmath221 is an outward normal of the side @xmath222;$ ] see fig .
[ hessactionfig ] . then @xmath223 given two linearly independent vectors @xmath229 we denote by @xmath230 the operator of projection onto @xmath231 along the vector @xmath232 that is , @xmath233 for @xmath234 and @xmath235 by _
cramer s rule _ , @xmath236 and hence @xmath237 which implies @xmath238 thus , is equivalent to @xmath239 it can be easily verified that @xmath240 therefore , implies @xmath241 we have @xmath242 [ u1u3eigv ] let @xmath1 be a strictly convex and @xmath43 regular body in @xmath243 let @xmath244 be distinct vectors from @xmath245 such that @xmath246 for @xmath247 and the segment @xmath194 $ ] is not an affine diameter of @xmath44 then the matrix @xmath248 has two distinct real eigenvalues .
furthermore , if @xmath249 are distinct eigenvectors of @xmath250 satisfying @xmath251 then @xmath252 the assumptions @xmath253 for @xmath247 imply that @xmath254 for @xmath255 by applied at @xmath65 and @xmath256 we get @xmath257 from and we see that @xmath100 is orthogonal to both @xmath258 and @xmath259 then @xmath258 and @xmath260 are parallel , which implies that @xmath261 is parallel to @xmath262 thus , @xmath102 is an eigenvector of the matrix @xmath263 analogous arguments show that also @xmath100 is an eigenvector of @xmath264 we show that it is not possible that all vectors from @xmath265 are eigenvectors of @xmath264 we introduce the centrally symmetric hexagon @xmath266 after , possibly , interchanging the roles of @xmath65 and @xmath256 we assume that the points @xmath267 are in counterclockwise order on @xmath117 ; see fig .
[ hexnormdetermfig ] .
( 95,60 ) ( 1,7 ) ( 63.5,32)@xmath225 ( 57,5)@xmath224 ( 66,52)@xmath75 ( -2,43)@xmath272 ( 92,42)@xmath273 ( 51,50.5)@xmath274 ( 81,30.5)@xmath275 ( 25,53)@xmath276 ( 26,16)@xmath277 ( 46,32)@xmath278 ( 79,53)@xmath279 ( 77,18)@xmath280 let @xmath225 be an outward normal of the side @xmath281 $ ] of @xmath282 let @xmath216 and @xmath283 be quadrilaterals constructed as in the statement of lemma [ hessquadinterp ] . by the choice of @xmath284 we have @xmath285 @xmath286
while @xmath277 follows @xmath287 and @xmath279 follows @xmath288 in counterclockwise order on @xmath215 consequently , @xmath289 $ ] is a common diagonal of @xmath216 and @xmath290 while the vertices @xmath274 and @xmath275 of @xmath283 lie in the relative interiors of the sides @xmath291 $ ] and @xmath292,$ ] respectively , of @xmath293 see fig .
[ hexnormdetermfig2 ] .
the latter implies that the diagonals @xmath294 $ ] and @xmath295 $ ] of the quadrilaterals @xmath216 and @xmath290 respectively , are not parallel .
hence , by lemma [ hessquadinterp ] , @xmath296 and @xmath297 are not parallel , which implies that the chosen @xmath225 is not an eigenvector of @xmath264 consequently , @xmath298 and @xmath299 are two distinct eigenspaces of @xmath300 and we arrive at the assertion .
[ hexdetermprop ] let @xmath1 be a strictly convex and @xmath43 regular body in @xmath24 let @xmath284 be a centrally symmetric convex hexagon with consecutive vertices @xmath301 in counterclockwise order . for @xmath302 we introduce the vectors @xmath303 ; see fig .
[ hexdetermnewfig ] .
then a translate of @xmath284 is inscribed in @xmath1 if and only if @xmath304 for every @xmath302 and @xmath305 let us show the necessity . since conditions and are invariant with respect to translations of @xmath315 we can assume that @xmath284 itself is inscribed in @xmath44 from the definition of @xmath316 and @xmath317 it follows that @xmath318 where @xmath126 thus , follows directly form and the definition of the function @xmath319 let us obtain .
by we have @xmath320 where @xmath321 the determinants @xmath322 and @xmath323 are strictly positive ; see .
consequently @xmath324 from we get the equalities @xmath325 and @xmath326 and by this also the equalities @xmath327 for @xmath328 hence we see that @xmath329 and therefore @xmath330 now let us show the sufficiency by contradiction .
assume that for some @xmath1 and @xmath315 satisfying the assumptions of the proposition , conditions and are fulfilled but no translate of @xmath284 is inscribed in @xmath44 by we see that for every @xmath331 the parallelogram @xmath332 is a translate of @xmath333 and , moreover , one has @xmath334 with appropriate @xmath335 if for some @xmath336 with @xmath337 the parallelograms @xmath338 @xmath339 share a diagonal , it follows that @xmath284 is a translate of @xmath340 a contradiction .
now we consider the case when no two distinct parallelograms @xmath332 and @xmath339 share a diagonal .
let @xmath126 in view of , we get that @xmath341 thus the diagonals @xmath342 $ ] and @xmath343 $ ] of @xmath344 and @xmath338 respectively , are translates of each other . by the assumption , these diagonals are distinct . thus ,
@xmath342 $ ] and @xmath343 $ ] are distinct chords of @xmath1 which are translates of @xmath345;$ ] see fig . [ threehexagons ] .
( 48,44 ) ( 3,4 ) ( 32,43)@xmath346 ( -3,26)@xmath347 ( 2,5)@xmath348 ( 45.5,23)@xmath349 ( -4.3,20)@xmath350 ( 30,5.8)@xmath351 ( 44,18)@xmath352 ( 3.7,32)@xmath353 ( 40,10)@xmath354 ( 41.8,40)@xmath355 ( 12,36)@xmath356 ( 12,1.5)@xmath357 first we show that there exists @xmath363 with @xmath364 such that @xmath253 for @xmath255 let @xmath187 be the center of @xmath167 and @xmath365 be the reflection of @xmath1 with respect to @xmath366 see fig .
[ hexnoaffdiag ] . assume first that @xmath367 then , in view of and , @xmath368 therefore @xmath117 and @xmath369 intersect transversally at @xmath370 for every @xmath144 moreover , either a small subarc of @xmath371_k$ ] with endpoint @xmath119 is contained in @xmath365 and a small subarc of @xmath371_k$ ] with endpoint @xmath270 is contained in @xmath372 or vice versa ( that is , a small subarc of @xmath371_k$ ] with endpoint @xmath119 is contained in @xmath372 and a small subarc of @xmath371_k$ ] with endpoint @xmath270 is contained in @xmath365 ) .
consequently , the arcs @xmath371_k$ ] and @xmath371_{k_c}$ ] intersect at some point @xmath373 distinct from @xmath119 and @xmath374 we define @xmath375 by construction , @xmath376 and @xmath377 are consecutive vertices of @xmath378 ; see fig .
[ hexnoaffdiag ] .
therefore @xmath192 satisfies the desired conditions . in the case
@xmath382 we can use similar arguments showing that the arcs @xmath383_k$ ] and @xmath383_{k_c}$ ] intersect at some point @xmath373 distinct from @xmath270 and @xmath384 thus , for that case we can define @xmath385 now let @xmath38 be a strictly convex and @xmath43 regular planar convex body with the same covariogram as @xmath44 by proposition [ hexdetermprop ] , a translate of @xmath386 is inscribed in @xmath387 without loss of generality we assume @xmath284 itself is inscribed in @xmath388 that is , @xmath389 and @xmath390 notice that the inequality @xmath391 implies that @xmath194 $ ] is not an affine diameter of @xmath1 or @xmath387 then , by lemma [ u1u3eigv ] , we have @xmath392 and we are done .
now assume that @xmath1 is not centrally symmetric .
then , by theorem [ central symmetry ] , there exists @xmath394 such that @xmath395 this implies @xmath396 let @xmath397 and @xmath398 be disjoint open neighborhoods of @xmath399 and @xmath400 respectively . in view of theorem [ frst : drv : cov : thm ] ( part [ ptrans : prop ] ) and proposition [ u1u3:determ ] , replacing @xmath38 by an appropriate translation or reflection , we can assume that @xmath401 and @xmath402 for @xmath255 let @xmath403 be a continuous , counterclockwise parametrization of a small boundary arc of @xmath1 such that @xmath404 and , for @xmath405 , one has @xmath406 , @xmath407 and @xmath408 for every @xmath409 .
we show by contradiction that for every @xmath409 the equalities @xmath410 are fulfilled .
assume the contrary .
then , by proposition [ u1u3:determ ] , there exists @xmath411 with @xmath412 such that @xmath413 and @xmath414 in particular we have @xmath415 .
since @xmath416 and since @xmath397 and @xmath398 are disjoint , there exists @xmath417 with @xmath418 such that @xmath419 lies outside @xmath420 hence @xmath421 but , by construction , we have @xmath422 a contradiction to proposition [ u1u3:determ ] .
the definition of @xmath185 implies @xmath423 , for each @xmath207 $ ] , and therefore , it also implies @xmath424 .
hence , in view of , we get @xmath425 consequently , by the strict convexity of @xmath38 , we also have @xmath426 .
the latter implies @xmath427_k = \bigl[p_4\bigl(l , x(0)\bigr),p_4\bigl(l , x(1)\bigr)\bigr]_l,\ ] ] and concludes the proof .
a. bjrner , m. las vergnas , b. sturmfels , n. white , and g. m. ziegler , _ oriented matroids _ ,
second ed . , encyclopedia of mathematics and its applications , vol .
46 , cambridge university press , cambridge , 1999 .
, _ le covariogramme gometrique des compacts convexes des @xmath430 _ , technical report n-2/86/g , centre de gostatistique , ecole nationale suprieure des mines de paris .
available at http://cg.ensmp.fr / bibliotheque/1986/matheron / rapport / doc_00258/mathero% n_rapport_00258.pdf[http://cg.ensmp.fr / bibliotheque/1986/matheron / rapport / doc_00258/mathero% n_rapport_00258.pdf ] , 1986 . | the covariogram @xmath0 of a convex body @xmath1 in @xmath2 is the function which associates to each @xmath3 the volume of the intersection of @xmath1 with @xmath4 . in 1986
g. matheron conjectured that for @xmath5 the covariogram @xmath0 determines @xmath1 within the class of all planar convex bodies , up to translations and reflections in a point .
this problem is equivalent to some problems in stochastic geometry and probability as well as to a particular case of the phase retrieval problem in fourier analysis .
it is also relevant for the inverse problem of determining the atomic structure of a quasicrystal from its x - ray diffraction image . in this paper
we confirm matheron s conjecture completely .
primary 60d05 ; secondary 52a10 , 52a22 , 52a38 , 42b10 autocorrelation , covariogram , cut - and - project scheme , geometric tomography , image analysis , phase retrieval , quasicrystal , set covariance |
in order to probe @xmath6 , @xmath5 decays are very promising .
there are already data on these modes available , which triggered a lot of theoretical work . from 1997 until 2000 , cleo reported results on @xmath0-averaged branching ratios @xmath10 ; in 1999 , also studies of @xmath0 asymmetries were reported @xcite . in the summer of 2000 ,
babar and belle announced their first @xmath5 branching ratios @xcite .
these results are collected in table [ tab : bpik ] . to get more familiar with @xmath5 modes ,
let us consider @xmath11 .
this channel receives contributions from penguin and colour - allowed tree - diagram - like topologies , as can be seen in fig .
[ fig : bpik - neutral ] .
because of the small ratio @xmath12 , the qcd penguin topologies dominate this decay , despite their loop suppression .
this interesting feature applies to all @xmath5 modes . because of the large top - quark mass
, we have also to care about ew penguins
. however , in the case of @xmath11 and @xmath13 , these topologies contribute only in colour - suppressed form and are hence expected to play a minor role . on the other hand ,
ew penguins contribute also in colour - allowed form to @xmath14 and @xmath15 , and may here even compete with tree - diagram - like topologies . because of the penguin dominance , @xmath5 modes represent sensitive probes for new - physics effects @xcite .
= 4.0truecm = 3.5truecm in the standard model , the @xmath16 isospin symmetry implies @xmath17 @xmath18 @xmath19\propto \left[e^{i\gamma}+q_{\rm ew}\right],\ ] ] where the @xmath20 and @xmath21 amplitudes are due to @xmath22colour - allowed + colour - suppressed@xmath23 tree - diagram - like and ew penguin topologies , respectively . a relation with an analogous phase structure holds also for the `` mixed '' @xmath24 , @xmath25 system .
so far , strategies to probe @xmath6 have focused on the following systems : @xmath26 , @xmath27 ( `` mixed '' ) @xcite@xcite , and @xmath28 , @xmath27 ( `` charged '' ) @xcite .
recently , it was pointed out that also the neutral combination @xmath26 , @xmath29 is very promising @xcite .
interestingly , already @xmath0-averaged @xmath5 branching ratios may lead to non - trivial constraints on @xmath6 , which rely on flavour - symmetry arguments , involving either @xmath16 or @xmath30 @xcite , and dynamical assumptions , concerning mainly the smallness of certain rescattering effects @xcite .
an example is @xmath31 .
there is still no theoretical consensus on the importance of such final - state interaction ( fsi ) processes , although , for instance , the `` qcd factorization '' approach @xcite does not suggest large effects . however , there are also experimental indicators for large fsi effects , for example the decays @xmath32 and @xmath33 , and methods to include them in the strategies to probe @xmath6 @xcite .
so far , these decays have not been observed and the rather strong experimental upper bounds on their @xmath0-averaged branching ratios are not in favour of dramatic fsi effects @xcite , as advocated in some of the papers given in ref .
let us therefore neglect these effects for a moment .
let us illustrate the derivation of the constraints on @xmath6 by considering the `` mixed '' @xmath24 , @xmath25 system @xcite .
if we neglect annihilation topologies , penguins with internal up quarks and colour - suppressed ew penguins , which are expected to play a very minor role ( but may be enhanced by large fsi effects ; see subsection [ sec : bpik - gen ] ) , we may write @xmath34 \,\equiv\ , -\left[-|p|e^{i\delta_p}+|t|e^{i\delta_t}e^{i\gamma } \right].\end{aligned}\ ] ] here @xmath35 and @xmath36 describe the penguin and tree - diagram - like topologies shown in fig .
[ fig : bpik - neutral ] , and @xmath37 and @xmath38 are @xmath0-conserving strong phases .
it is convenient to re - write the @xmath11 decay amplitude as @xmath39,\ ] ] where @xmath40 and @xmath41 .
consequently , we obtain the following expressions for the @xmath0-averaged decay amplitudes : latexmath:[\[\begin{aligned } \left\langle|a(b^\pm\to\pi^\pm k)|^2\right\rangle&=&|\tilde p|^2\\ \left\langle|a(b_d\to\pi^\mp k^\pm)|^2\right\rangle&= & the @xmath16 isospin symmetry implies @xmath43 , , which corresponds to ( [ bpik - ampl ] ) . ]
we arrive at @xmath44 keeping now the hadronic quantities @xmath45 and @xmath46 as `` free '' parameters yields @xmath47 which implies @xmath48 @xmath49 @xmath50 @xcite , with @xmath51 this constraint on @xmath6 is only effective if @xmath52 is found to be smaller than one . in 1997 ,
when cleo reported the first result on the @xmath0-averaged @xmath24 , @xmath25 branching ratios , the result was @xmath53 .
the central value @xmath54 would imply @xmath55 , thereby excluding a large range in the @xmath56@xmath57 plane . using the present cleo data @xcite
, we obtain @xmath58 .
unfortunately , the present experimental uncertainties are too large to draw any conclusions and to decide whether @xmath59 . if the parameter @xmath46 is fixed , for example through `` factorization ''
@xcite , stronger constraints on @xmath6 can be obtained , which are also effective for @xmath60 @xcite . in order to constrain @xmath6 ,
the key quantities are the following ratios : @xmath61 where we have also taken into account the cleo results reported in @xcite .
if we employ the @xmath16 flavour symmetry and certain dynamical assumptions , concerning mainly the smallness of fsi effects , we may derive a general parametrization for ( [ cleo - mix])([cleo - neut ] ) ( for an explicit example , see ( [ r - param ] ) ) @xcite , @xmath62 where @xmath63 denotes the ratio of ew penguins to `` trees '' , @xmath64 is the ratio of `` trees '' to qcd penguins , and @xmath65 is the @xmath0-conserving strong phase between `` tree '' and qcd penguin amplitudes .
the @xmath63 can be fixed through theoretical arguments : in the `` mixed '' @xmath5 system , we have @xmath66 , as ew penguins contribute only in colour - suppressed form ; in the charged @xcite and neutral @xcite @xmath5 systems , @xmath67 and @xmath68 can be fixed through the @xmath30 flavour symmetry .
on the other hand , the @xmath64 can be determined with the help of additional experimental information : in the `` mixed '' system , @xmath46 can be fixed , for example , through arguments based on `` factorization '' @xcite , whereas @xmath69 and @xmath70 can be determined from @xmath71 by using again the @xmath30 flavour symmetry @xcite . at this point , a comment on the fsi effects discussed in subsection [ sec : bpik - gen ] is in order . whereas the determination of @xmath72 and @xmath46 as sketched above may be affected by fsi effects ,
this is _ not _ the case for @xmath73 and @xmath74 , since here @xmath30 suffices .
nevertheless , we have to assume that @xmath13 or @xmath75 do _ not _ involve a @xmath0-violating weak phase , i.e. @xmath76 this relation may be affected by rescattering processes as follows : @xmath77,\ ] ] where @xmath78 is doubly cabibbo - suppressed and is naively expected to be negligibly small . in the `` qcd factorization '' approach @xcite , there is no significant enhancement of @xmath78 through rescattering processes .
however , as we have already noted , there is still no theoretical consensus on the importance of fsi effects . in the charged @xmath5 strategy to probe @xmath6 ,
they can be taken into account through @xmath30 flavour - symmetry arguments and additional data on @xmath79 decays .
on the other hand , in the case of the neutral strategy , fsi effects can be included in an _ exact manner _ with the help of the mixing - induced @xmath0 asymmetry @xmath80 @xcite .
in contrast to @xmath63 and @xmath64 , the strong phase @xmath65 suffers from large hadronic uncertainties and is essentially unknown .
however , we can get rid of @xmath65 by keeping it as a `` free '' variable , yielding minimal and maximal values for @xmath81 : @xmath82 keeping in addition @xmath64 as a free variable , we obtain another less restrictive minimal value for @xmath81 : @xmath83 in fig .
[ fig : rn ] , we show the dependence of ( [ const1 ] ) and ( [ const2 ] ) on @xmath6 for the neutral @xmath5 system .
curves look very similar . ] here the crossed region below the @xmath84 curve , which is described by ( [ const2 ] ) , is excluded . on the other hand ,
the shaded region is the allowed range ( [ const1 ] ) for @xmath85 , arising in the case of @xmath86 .
[ fig : rn ] allows us to read off immediately the allowed region for @xmath6 for a given value of @xmath85 . using the central value of the present cleo result ( [ cleo - neut ] ) , @xmath87
, the @xmath84 curve implies @xmath88 .
the corresponding situation in the @xmath89@xmath57 plane is shown in fig .
[ fig : rho - eta - bounds ] , where the crossed region is excluded and the circles correspond to @xmath90 .
as the theoretical expression for @xmath68 is proportional to @xmath91 , the constraints in the @xmath89@xmath57 plane are actually more appropriate than the constraints on @xmath6 .
if we use additional information on @xmath70 , we may put even stronger constraints on @xmath6 . for @xmath86
, we obtain the allowed range @xmath92 .
it is interesting to note that the @xmath84 curve is only effective for @xmath93 , which is favoured by the most recent cleo data @xcite .
a similar pattern is also exhibited by the first belle results @xcite , yielding @xmath94 .
for the central value @xmath95 , ( [ const2 ] ) is not effective and @xmath69 has to be fixed to constrain @xmath6 . using @xmath96 , we obtain @xmath97 .
although it is too early to draw definite conclusions , it should be emphasized that the present cleo results on @xmath81 prefer the second quadrant for @xmath6 , i.e. @xmath98 .
similar conclusions were also obtained using other @xmath5 , @xmath99 strategies @xcite .
interestingly , such a situation would be in conflict with the standard analysis of the unitarity triangle @xcite , yielding @xmath100 .
the @xmath81 allow us to determine @xmath101 as functions of @xmath6 , thereby providing also constraints on the strong phases @xmath65 @xcite .
interestingly , the present cleo data are in favour of @xmath102 , which would be in conflict with `` factorization '' .
moreover , they point towards a positive value of @xmath103 , which would be in conflict with the theoretical expectation of equal signs for @xmath103 and @xmath104 .
if future data should confirm the `` puzzling '' situation for @xmath6 and @xmath105 , which is favoured by the present @xmath5 cleo data , it may be an indication for new - physics contributions to the ew penguin sector , or a manifestation of large flavour - symmetry - breaking effects . in order to distinguish between these possibilities ,
further studies are needed .
as soon as @xmath0 asymmetries in @xmath26 or @xmath28 are observed , @xmath106 we may even _ determine _ @xmath6 and @xmath65 .
here we may also arrive at a situation , where the @xmath5 observables do not provide any solution for these quantities , which would be an immediate indication for new physics @xcite .
calculations of @xmath107 modes are usually based on a perturbative picture , where strong phases are obtained from absorptive parts of loop diagrams .
this approach is referred to as the `` bander silverman soni ( bss ) mechanism '' @xcite .
recently , a conceptual improvement of this formalism was presented in ref .
@xcite ( see also @xcite ) , where it is argued that there is a heavy - quark expansion for non - leptonic @xmath1-decays into two light mesons , yielding transition amplitudes of the structure given in ( [ qcd - factor ] ) . in this approach , soft non - factorizable contributions and fsi effects are suppressed by @xmath108 .
however , the question arises whether the @xmath109-quark mass is large enough to suppress these terms sufficiently . moreover , there are problems due to `` chirally enhanced '' terms , which are formally suppressed by @xmath110 , but are numerically of @xmath111 .
another `` perturbative '' qcd approach to deal with non - leptonic charmless @xmath1 decays was developed in ref .
. many calculations of @xmath5 , @xmath99 can be found in the literature ( see , for instance , @xcite ) .
the results for the @xmath0-averaged branching ratios are generally in rather good agreement with the cleo data
. however , there are two exceptions : @xmath7 and @xmath112 . concerning the former decay , the calculations favour a value at the @xmath113 level , whereas cleo finds br@xmath114 .
due to interference between tree - diagram - like and penguin topologies , the theoretical predictions depend on @xmath6 ; the cleo result would be in favour of @xmath115 @xcite .
on the other hand , the first babar result is br@xmath116 . concerning @xmath112 , the calculations favour a value at the @xmath117 level , which is essentially independent of @xmath6 and smaller than the cleo result br@xmath118 .
interestingly , the first belle result @xmath119 is also in favour of a large @xmath0-averaged @xmath29 branching ratio . because of the large experimental uncertainties
, we can not yet draw any definite conclusions .
however , the experimental situation should improve significantly in the future .
as we have seen in subsection [ bpipi - alpha ] , @xmath7 is usually considered as a tool to determine @xmath120 .
unfortunately , the extraction of @xmath121 from @xmath122 is affected by large penguin uncertainties , and the strategies to control them through additional data are challenging . in this section , we discuss a new approach to employ @xmath7 : combining this mode with @xmath8
its @xmath123-spin - spin is a subgroup of @xmath124 , relating down and strange quarks to each other . ]
counterpart a simultaneous determination of @xmath125 and @xmath6 becomes possible @xcite . in this method ,
no model - dependent assumptions are required , and fsi effects , which led to considerable attention in the @xmath5 strategies to probe @xmath6 , do not lead to any problems .
the theoretical accuracy is hence only limited by @xmath123-spin - breaking effects .
this approach is promising for cdf - ii ( @xmath126 ) @xcite , and ideally suited for lhcb ( @xmath127 ) @xcite .
there are also other interesting strategies to determine @xmath6 with the help of the @xmath123-spin symmetry : @xmath128 , @xmath129 @xcite or @xmath130 @xcite , as well as @xmath131 @xcite . for
the `` prehistory '' of @xmath123-spin arguments in @xmath1 decays , the reader is referred to ref .
@xcite .
looking at the diagrams shown in fig .
[ fig : bpipi ] , we observe that @xmath8 is obtained from @xmath7 by interchanging all down and strange quarks . the structure of the corresponding decay amplitudes is given as follows : @xmath132\ ] ] @xmath133,\ ] ] where @xmath134 and @xmath135 take the same form as @xmath136 and @xmath137 ( see ( [ c - def ] ) and ( [ d - def ] ) ) . using the formalism discussed in subsection [ subsec : cpasym ] yields @xmath138 @xmath139 where the standard - model expectation @xmath140 can be probed through the decay @xmath141 .
consequently , we have four observables , depending on six `` unknowns '' .
however , since @xmath7 and @xmath8 are related to each other by interchanging all down and strange quarks , the @xmath123-spin flavour symmetry of strong interactions implies @xmath142 using this relation , the four observables ( [ asym-1])([asym-4 ] ) depend on the four quantities @xmath143 , @xmath144 , @xmath125 and @xmath6 , which can hence be determined . the @xmath123-spin arguments can be `` minimized '' , if we employ @xmath125 as an input , which can be determined straightforwardly from @xmath145 .
the observables @xmath146 and @xmath122 allow us then to eliminate the strong phase @xmath144 and to determine @xmath143 as a function of @xmath6 .
analogously , @xmath147 and @xmath148 allow us to eliminate the strong phase @xmath149 and to determine @xmath150 as a function of @xmath6 .
the corresponding contours in the @xmath6@xmath143 and @xmath6@xmath150 planes can be fixed in a _ theoretically clean _ way . using the @xmath123-spin relation @xmath151 ,
these contours allow the determination both of the ckm angle @xmath6 and of the hadronic quantities @xmath143 , @xmath144 , @xmath149 .
let us illustrate this approach by considering the following example : @xmath152 which corresponds to the input parameters @xmath153 , @xmath154 , @xmath155 , @xmath156 and @xmath157 . in fig .
[ fig : bskkcont ] , the corresponding contours in the @xmath6@xmath143 and @xmath6@xmath150 planes are represented by the solid and dot - dashed lines , respectively .
their intersection yields a twofold solution for @xmath6 , given by @xmath158 and our input value of @xmath159 .
the dotted line is related to the quantity @xmath160,\ ] ] which can be combined with @xmath148 through the @xmath123-spin relation ( [ u - spin - rel ] ) to fix another contour in the @xmath6@xmath143 plane .
combining all contours in fig .
[ fig : bskkcont ] with one another , we obtain a single solution for @xmath6 , which is given by the `` true '' value of @xmath159 . as
an interesting by - product , the penguin parameters @xmath143 and @xmath144 , @xmath149 can be determined as well , allowing a comparison with theoretical predictions and valuable insights into hadronic physics .
it should be emphasized that the theoretical accuracy of @xmath6 and the hadronic parameters @xmath143 , @xmath144 , @xmath149 is only limited by @xmath123-spin - breaking effects . in particular ,
it is not affected by fsi or penguin effects .
interestingly , the relation ( [ u - spin - rel ] ) does not receive @xmath123-spin - breaking corrections within a modernized version of the `` bander silverman soni mechanism '' @xcite , making use among other things of the factorization hypothesis to estimate the relevant hadronic matrix elements @xcite .
this remarkable feature strengthens our confidence into the @xmath123-spin relations used for the extraction of @xmath9 and @xmath6
. however , further theoretical studies of the @xmath123-spin - breaking effects in the @xmath7 , @xmath8 system , employing , for example , the `` qcd factorization '' approach @xcite , would be desirable .
apart from these theoretical considerations , it is also possible to obtain experimental insights into @xmath123-spin breaking .
a first consistency check is provided by @xmath161 .
moreover , we may determine the normalization factors @xmath162 and @xmath163 of the @xmath164 and @xmath165 decay amplitudes ( see ( [ bdpipi - ampl0 ] ) and ( [ bskk - ampl0 ] ) ) with the help of the corresponding @xmath0-averaged branching ratios . comparing them with the `` factorized '' result @xmath166 \left[\frac{f_{b_sk}(m_k^2;0^+)}{f_{b_d\pi}(m_\pi^2;0^+)}\right ] \left(\frac{m_{b_s}^2-m_k^2}{m_{b_d}^2-m_\pi^2}\right),\ ] ] we have another interesting probe for @xmath123-spin - breaking effects .
moreover , the @xmath123-spin relation ( [ u - spin - rel ] ) also implies @xmath167= -\left|\frac{{\cal c}'}{{\cal c}}\right|^2\times \left[\frac{\mbox{br}(b_d\to\pi^+\pi^-)}{\mbox{br}(b_s\to k^+k^-)}\right].\ ] ] in order to obtain further insights , the @xmath168 , @xmath169 system would be of particular interest , allowing us to determine @xmath6 together with the mixing phases @xmath170 and @xmath171 , and tests of several @xmath123-spin relations @xcite . further strategies to explore @xmath123-spin - breaking effects were recently discussed in ref .
@xcite .
since penguin processes play an important role in the decays @xmath172 and @xmath7 , they and the strategy to determine @xmath6 , where moreover the unitarity of the ckm matrix is employed may well be affected by new physics .
interestingly , the standard model implies a rather restricted region in the space of the @xmath0-violating observables of the @xmath8 , @xmath7 system @xcite .
a future measurement of observables lying significantly outside of this allowed region would be an immediate indication for new physics .
on the other hand , if the observables should lie within the region predicted by the standard model , we can extract a value for @xmath6 by following the strategy discussed above .
this value for @xmath6 may well be in conflict with other approaches , which would then also indicate the presence of new physics .
let us finally make a few remarks on other `` rare '' @xmath1 decays , which occur only at the one - loop level in the standard model , and involve @xmath173 or @xmath174 flavour - changing neutral - current transitions : @xmath175 , @xmath176 , @xmath177 , inclusive decays , etc . within the standard model , these transitions exhibit small branching ratios at the @xmath178 level , do not show sizeable @xmath0-violating effects , and depend on @xmath179 and @xmath180 .
a measurement of these ckm factors from rare decays would be complementary to the one from @xmath181@xmath182 mixing .
since they are absent at the tree level in the standard model , rare @xmath1-decays represent interesting probes to search for new physics , and have many other interesting aspects . for detailed discussions ,
the reader is referred to the reviews listed in refs .
the phenomenology of non - leptonic decays of @xmath1-mesons is very rich and has been a field of very active research over the last couple of years .
these modes provide a very fertile testing ground for the standard - model description of @xmath0 violation , and allow the direct determination of the angles of the unitarity triangles . here
the goal is to overconstrain these triangles , which may open a window to the physics beyond the standard model . as by - products ,
some strategies allow also the determination of interesting hadronic parameters and provide valuable insights into hadronic physics . moreover , there are many other exciting aspects , for instance the physics potential of certain rare @xmath1-decays . in view of the rich experimental @xmath1-physics programme of this decade and
the strong interaction between theory and experiment , i have no doubt that an exciting future is ahead of us .
i would like to thank the organizers of this nato advanced study institute for inviting me to this very interesting meeting in such a pleasant environment .
special thanks go to gustavo branco , gui rebelo and juca silva
marcos for their splendid hospitality . for reviews , see nir , y. ( 1999 ) , iassns - hep-99 - 96 [ hep - ph/9911321 ] ; + rosner , j.l .
( 2000 ) , efi-2000 - 16 [ hep - ph/0005258 ] ; + gronau , m. ( 1999 ) , slac - pub-8221 [ hep - ph/9908343 ] ; + fleischer , r. ( 2000 ) , _ nucl .
instrum .
meth . _ * a446 * , 1 .
branco , g. , lavoura , l. and silva , j. ( 1999 ) , _
@xmath0 violation _ ,
oxford science publications , clarendon press , oxford ; + bigi , i.i . and
sanda , a.i .
( 2000 ) , _
@xmath0 violation _ , cambridge monographs on particle physics , nuclear physics and cosmology , cambridge university press , cambridge . for reviews ,
see grossman , y. , nir , y. and rattazzi , r. ( 1997 ) , in _ heavy flavours ii _ , p. 755 , eds .
buras , a. and lindner , m. , world scientific , singapore [ hep - ph/9701231 ] ; + gronau , m. and london , d. ( 1997 ) , _ phys .
. _ * d55 * , 2845 ; + nir , y. and quinn , h.r .
( 1992 ) , _ annu .
nucl . part .
_ * 42 * , 211 ; + fleischer , r. ( 1997 ) , in the proceedings of the 7th international symposium on heavy flavor physics , santa barbara , california , july 711 , 1997 , ed .
campagnari , c. , world scientific , singapore , p. 155
[ hep - ph/9709291 ] ; + wolfenstein , l. ( 1998 ) , _ phys .
_ * d57 * , 6857 . , eds .
harrison , p. and quinn , h. ( 1998 ) , slac report 504 ; + workshop on @xmath1 physics at the tevatron run ii and beyond , please have a look at http://www-theory.fnal.gov/people/ligeti/brun2/ ; + report of the @xmath109-decay working group of the workshop _ standard model physics ( and more ) at the lhc _ , ball , p. _ et al . _ ( 2000 ) , cern - th/2000 - 101 [ hep - ph/0003238 ]
. glashow , s.l .
( 1961 ) , _ nucl . phys .
_ * 22 * , 579 ; + weinberg , s. ( 1967 ) , _ phys .
_ * 19 * , 1264 ; + salam , a. ( 1968 ) , in _ elementary particle theory _
svartholm , n. , almqvist and wiksell , stockholm .
buras , a.j .
( 1998 ) , in the proceedings of the les houches 1997 summer school on theoretical physics : probing the standard model of particle interactions , les houches , france , july 28 september 5 , 1997 , eds .
gupta , r. , morel , a. , de rafael , e. and david , f. , north - holland , amsterdam [ hep - ph/9806471 ] .
see , for instance , schwinger , j. ( 1964 ) , _ phys .
* 12 * , 630 ; + farikov , d. and stech , b. ( 1978 ) , _ nucl .
phys . _ * b133 * , 315 ; + cabibbo , n. and maiani , l. ( 1978 ) , _ phys .
lett . _ * b73 * , 418 [ e : ibid . *
b76 * , 663 ] ; + bjorken , j.d .
( 1989 ) , _ nucl .
suppl . ) _ * b11 * , 325 ; + dugan , m. and grinstein , b. ( 1991 ) , _ phys .
lett . _ * b255 * , 583 ; + politzer , h.d . and wise , m.b .
( 1991 ) , _ phys .
_ * b257 * , 399 ; + neubert , m. and stech , b. ( 1997 ) , in _ heavy flavours ii _ , p. 294 , eds .
buras , a.j . and lindner , m. , world scientific , singapore [ hep - ph/9705292 ] .
kagan , a.l . and
neubert , m. ( 2000 ) , _ phys .
_ * b492 * , 115 ; + silva , j.p . and wolfenstein , l. ( 2000 ) , slac - pub-8548 [ hep - ph/0008004 ] ; + eyal , g. , nir , y. and perez , g. ( 2000 ) , _ j. high energy phys .
_ * 0008 * , 028 ; + xing , z .- z .
( 2000 ) , hep - ph/0008018 ; + buras , a.j . and buras , r. ( 2000 ) , tum - hep-285 - 00 [ hep - ph/0008273 ] .
gronau , m. ( 1993 ) , _ phys .
_ * b300 * , 163 ; + silva , j.p . and wolfenstein , l. ( 1994 ) , _ phys
. rev . _ * d49 * , r1151 ; + aleksan , r. _ et al . _
( 1995 ) , _ phys .
lett . _ * b356 * , 95 ; + ward , b.f .
( 1995 ) , _ phys . rev .
_ * d51 * , 6253 ; + dejongh , f. and sphicas , p. ( 1996 ) , _ phys . rev .
_ * d53 * , 4930 ; + ciuchini , m. _ et al . _
( 1997 ) , _ nucl .
phys . _ * b501 * , 271 ; + marrocchesi , p.s . and paver , n. ( 1998 ) , _ int .
_ * a13 * , 251 ; + ali , a. , kramer , g. and l , c .- d . ( 1999 ) , _ phys . rev .
* d59 * , 014005 . fleischer , r. and mannel , t. ( 1997 ) , ttp-97 - 22 [ hep - ph/9706261 ] ; + choudhury , d. , dutta , b. and kundu , a. ( 1999 ) , _ phys . lett .
_ * b456 * , 185 ; + he , x .- g . , hsueh , c .-
l . and shi , j .- q .
( 2000 ) , _ phys .
lett . _ * 84 * , 18 ; + grossman , y. , neubert , m. and kagan , a.l . ( 1999 ) , _ j. high energy phys .
_ * 9910 * , 029 .
hernndez , o. , london , d. , gronau , m. and rosner , j. ( 1994 ) , _ phys .
. _ * b333 * , 500 ; _ phys . rev . _ * d50 * , 4529 ; + gronau , m. , hernandez , o. , london , d. and rosner , j. ( 1995 ) , _ phys . rev .
_ * d52 * , 6356 .
wolfenstein , l. ( 1995 ) , _ phys .
rev . _ * d52 * , 537 ; + donoghue , j.f . ,
golowich , e. , petrov , a.a . and soares , j.m .
( 1996 ) , _ phys .
lett . _ * 77 * ,
2178 ; + buras , a. , fleischer , r. and mannel , t. ( 1998 ) , _ nucl . phys .
_ * b533 * , 3 ; + neubert , m. ( 1998 ) , _ phys .
_ * b424 * , 152 ; + grard , j .-
m . and weyers , j. ( 1999 ) , _ eur . phys . j. _ * c7 * , 1 ; + falk , a. , kagan , a. , nir , y. and petrov , a. ( 1998 ) , _ phys . rev .
_ * d57 * , 4290 ( 1998 ) ; + atwood , d. and soni , a. ( 1998 ) , _ phys .
_ * d58 * , 036005 .
fleischer , r. ( 1998 ) , _ phys .
. _ * b435 * , 221 ; + gronau , m. and rosner , j. ( 1998 ) , _ phys . rev . _ * d58 * , 113005 ; + gronau , m. and pirjol , d. ( 1999 ) , _ phys . lett . _ * b449 * , 321 ; + agashe , k. and deshpande , n. ( 1999 ) , _ phys .
_ * b451 * , 215 and * b454 * , 359 .
ali , a. , kramer , g. and l , c .- d .
( 1998 ) , _ phys .
rev . _ * d58 * , 094009 ; + keum , y .- y . ,
h - n . , and sanda , a.i .
( 2000 ) , ncku - hep-00 - 01a [ hep - ph/0004173 ] ; + muta , t. , sugamoto , a. , yang , m .- z . and yang , y .- d .
( 2000 ) , _ phys .
rev . _ * d62 * , 094020 .
dunietz , i. ( 1993 ) , in the proceedings of the workshop on @xmath1 physics at hadron accelerators , snowmass , colorado , eds .
mcbride , p. and shekhar mishra , c. , p. 83 ; + lipkin , h. ( 1997 ) , _ phys .
lett . _ * b415 * , 186 ; + buras , a. , fleischer , r. and mannel , t. ( 1998 ) , in ref .
@xcite ; + falk , a. , kagan , a. , nir , y. and petrov , a. ( 1998 ) , in ref .
@xcite ; + pirjol , d. ( 1999 ) , _ phys .
rev . _ * d60 * , 054020 .
ali , a. ( 1997 ) , in the proceedings of the 7th international symposium on heavy flavor physics , santa barbara , california , july 711 , 1997 , ed .
campagnari , c. , world scientific , singapore , p. 196
[ hep - ph/9709507 ] ; + greub , c. ( 1999 ) , butp-99 - 22 [ hep - ph/9911348 ] ; + misiak , m. ( 2000 ) , hep - ph/0002007 . | after an introduction to the standard - model description of @xmath0 violation and a brief look at the present status of this phenomenon in the kaon system , a classification of non - leptonic @xmath1-decays is given and the formalism of @xmath2@xmath3 mixing is discussed .
we then turn to the @xmath1-factory benchmark modes , @xmath0 violation in charged @xmath1 decays , and the @xmath4-meson system .
finally , we focus both on @xmath5 decays , which play an important role to probe the ckm angle @xmath6 , and on the @xmath7 , @xmath8 system , which allows an interesting determination of @xmath9 and @xmath6 .
[ cols= " < " , ] |
we consider the standard basis in the set of ( non analytic ) complex polynomials in @xmath0 : @xmath1 as usual @xmath2 is called the total degree of the given monomial . in what follows we use the following notations
: @xmath3 denotes the set of polynomials with complex coefficients of total degree less or equal @xmath4 and @xmath5 denotes the uniform norm of @xmath6 in the complex ball @xmath7 analogously to the classical chebyshev polynomial , we consider the best approximation on the ball @xmath8 of the monomial @xmath9 by polynomials of total degree less than @xmath10 such a polynomial @xmath11 which we call a polynomial of least deviation from zero on @xmath12 is not unique but the minimal deviation @xmath13 is well defined , @xmath14 it is convenient to work with the normalized polynomial @xmath15 thus @xmath16 where @xmath17 .
+ concerning polynomials of least deviation from zero on the unit ball in @xmath18 several approaches are known so far , see @xcite and also @xcite . of foremost importance to us
was the representation of the extremal polynomial given by bra in @xcite .
here we essentially simplify and generalize his construction . in approximation theory , a special role of the conformal mappings on so - called comb - like domains
is well known , see e.g. the book @xcite , the survey @xcite and the references on original papers therein , in particular @xcite . for recent developments in this direction ,
see @xcite . by analogy with the maclane - vinberg special representation for polynomials and entire functions @xcite , in this paper to nonnegative real numbers @xmath19 and an integer @xmath20
we associate a horizontal strip with @xmath21 horizontal cuts , see fig .
[ fig1 ] : @xmath22 note that the boundary of the domain contains @xmath23 infinite points : @xmath24 let @xmath25 be the conformal mapping of the upper half - plane onto @xmath26 with the following normalization @xmath27 it is easy to see that it has the following asymptotics at infinity ( @xmath28 ) @xmath29 the real constant @xmath30 is uniquely defined by the domain ( a kind of capacity ) . due to the evident symmetry @xmath31 we can assume that @xmath32 and @xmath33 .
our first result is [ th1 ] in the above introduced notations @xmath34 below we give the representation for a polynomial @xmath35 of least deviation from zero . for this
, we establish a connection between the conformal mapping @xmath36 and a weighted 1-d extremal problem on @xmath37 $ ] . in a sense , in the following proposition
we define _ uniform jacobi polynomials _
, compare to the classical orthogonal ones @xcite .
[ pr2 ] let @xmath38 , @xmath39 .
then @xmath40 where @xmath41 is the polynomial of least deviation from zero on @xmath37 $ ] with respect to the weight @xmath42
. moreover @xmath43 that is , @xmath44 as before @xmath45 .
we point out that @xmath46 for all @xmath47 .
[ th3 ] let @xmath48 and @xmath49 let us factorize @xmath50 in polynomials of degrees @xmath51 and @xmath52 respectively in the following way @xmath53 then @xmath54 finally we present the following asymptotic relation for the value of the minimal deviation . [ th4 ] assume that the following limits exist @xmath55 where @xmath56
. then @xmath57
first , we reduce our complex two - dimensional approximation problem to a weighted approximation problem in two real variables on the standard triangle @xmath58 . for a continuous function @xmath59 on @xmath60
we define @xmath61 by @xmath62 we denote a normalized polynomial of least deviation from zero on @xmath63 with respect to the weight function @xmath64 .
[ prop1 ] let @xmath65 and @xmath66 then @xmath67 is a normalized polynomial of least deviation from zero on @xmath8 , that is , @xmath68 let us remark that due to the symmetries of @xmath12 if @xmath35 is a polynomial of least deviation from zero , then for any @xmath69,$ ] the polynomials @xmath70 and @xmath71 are also polynomials of least deviation from zero .
it is easy to see that the polynomial in is of the form @xmath72 \\
= : \lambda_{k_1,l_1,k_2,l_2 } z_1^{k_1 - l_1 } z_2^{k_2 - l_2 } \tilde p_{l_1,l_2}(|z_1|^2,|z_2|^2 ) \end{split}\ ] ] where @xmath73 .
note that @xmath74 where @xmath75 is still a normalized polynomial of least deviation from zero on @xmath76 . since @xmath77 is equivalent to @xmath78 where @xmath79 @xmath80 we have that @xmath81 which gives the assertion .
now we give a sufficient condition for a polynomial to be a weighted polynomial of least deviation from zero on @xmath63 . in the next section
we show the existence of a polynomial satisfying this condition . as before @xmath82 denotes the normalized polynomial of least deviation from zero on @xmath37 $ ] with respect to the weight function @xmath83 .
[ prop2 ] let @xmath84 and let @xmath85 be the leading coefficient of @xmath82 , @xmath86 if there exists a polynomial @xmath87 such that @xmath88 for all @xmath89,$ ] then @xmath90 that is , the given @xmath91 is a normalized polynomial of least deviation from zero on @xmath63 with respect to the weight @xmath92 actually , we have to prove . by the fact that @xmath93 is a polynomial of least deviation from zero , from we have immediately that @xmath94 on the other hand , let us restrict @xmath93 to the line @xmath95 @xmath96 since @xmath97 for all @xmath98,$ ] the extremal property of @xmath99 implies @xmath100 thus the statement is proved .
let @xmath101 @xmath102 from the schwarz - christoffel formula , see e.g. @xcite , we obtain the following expression for the differential of the conformal mapping @xmath103 @xmath104 where @xmath105 is a constant . having in mind the asymptotic behavior at the infinite boundary points of the domain @xmath26 we get the following expansion into partial fraction for : @xmath106 hence @xmath107 where @xmath108 is a constant .
relation follows now immediately from . from the boundary correspondence for
the given conformal mapping we get that the function @xmath109 alternates @xmath21 times between @xmath110 on @xmath37 $ ] , see fig . [ fig2 ] .
thus the chebyshev alternation theorem implies that @xmath111 is indeed the polynomial of least deviation from zero with respect to the given weight with leading coefficient @xmath112 + by proposition [ prop1 ] , the assertion of theorem [ th3 ] follows if we are able to show that @xmath113 is a normalized polynomial of least deviation from zero on @xmath63 with respect to the weight @xmath114 for which we will use proposition [ prop2 ] . by restricting the polynomial @xmath91 to the line @xmath115 @xmath116
we obtain that for all @xmath89:$ ] @xmath117 thus it remains to show that @xmath118 .
let @xmath119 clearly @xmath120 is a product of two univariate functions , see .
we normalize the first factor @xmath121 by the condition @xmath122 . due to the definition of @xmath123 we have @xmath124 .
thus @xmath125 in addition , since @xmath126 we can easily check , see and , that @xmath127 is strictly increasing for @xmath128 $ ] , in particular @xmath129 here . since @xmath130 , @xmath131 is strictly increasing and @xmath132 for @xmath133 $ ] .
we note that by and : latexmath:[\[\label{eqint2 } for all @xmath89.$ ] in order to show the main claim @xmath135 we distinguish three regions in @xmath136 if @xmath137 $ ] then yields @xmath138 where the last inequality follows by the above listed properties of @xmath139 .
similarly we obtain @xmath140 if @xmath141 $ ] .
thus @xmath142 , t_2\in [ 0,1-\eta_{l_1}].\ ] ] if @xmath143 , then since @xmath144 is increasing on @xmath145,$ ] it follows that @xmath146 hence follows by .
if @xmath147 , then since @xmath139 is increasing on @xmath148,$ ] it follows that @xmath149 hence follows again by . by combining the three cases it follows that relation holds for all @xmath150 in conclusion , relations and being proved , by proposition [ prop2 ] it follows that @xmath151 is a normalized polynomial of least deviation from zero on @xmath63 with respect to the weight @xmath114 and hence , the polynomial given by is a polynomial of least deviation from zero on @xmath12 which also proves theorem [ th1 ] .
we need certain properties of the conformal mapping @xmath152 of the upper half - plane onto the domain @xmath153 see fig .
[ fig3 ] . due to the schwarz - christoffel formula @xcite , it is of the form @xmath154 where @xmath155 , @xmath156 , are the preimages of the angle - points @xmath157 and @xmath158 respectively . as before three `` infinite points '' in the domain
correspond to @xmath159 and @xmath160 and due to the size of corresponding strips we have the following relations @xmath161 as the sequence of domains @xmath168 converges to @xmath169 as @xmath170 it follows by carathodory s theorem , see e.g. @xcite , that for the sequence of conformal mappings it holds that @xmath171 therefore @xmath172 n. i. akhiezer and b. ya .
levin , _ generalization of s. n. bernstein s inequality for derivatives of entire functions_. ( russian ) 1960 issledovanija po sovremennym problemam teorii funkcii kompleksnogo peremennogo pp .
111165 gosudarstv .
fiz .- mat . lit . ,
moscow .
levin , _ classification of closed sets on @xmath176 and representation of a majorant_. iii .
( russian ) teor .
funktsii funktsional .
i prilozhen .
52 ( 1989 ) , 2133 ; translation in j. soviet math .
52 ( 1990 ) , no . 5 , 33643372 .
f. nazarov , f. peherstorfer , a. volberg , p. yuditskii , _ asymptotics of the best polynomial approximation of @xmath177 and of the best laurent polynomial approximation of @xmath178 on two symmetric intervals_. constr .
approx . 29
( 2009 ) , no .
1 , 2339 .
m. sodin and p. yuditskii , _ functions that deviate least from zero on closed subsets of the real axis_. ( russian ) algebra i analiz 4 ( 1992 ) , no .
2 , 161 ; translation in st .
petersburg math .
j. 4 ( 1993 ) , no .
2 , 201249 . | we consider the problem of finding a best uniform approximation to the standard monomial on the unit ball in @xmath0 by polynomials of lower degree with complex coefficients .
we reduce the problem to a one - dimensional weighted minimization problem on an interval . in a sense ,
the corresponding extremal polynomials are uniform counterparts of the classical orthogonal jacobi polynomials
. they can be represented by means of special conformal mappings on the so - called comb - like domains . in these terms , the value of the minimal deviation and the representation for a polynomial of best approximation for the original problem are given .
furthermore , we derive asymptotics for the minimal deviation . _ dedicated to the memory of franz peherstorfer _ |
it is our objective in the present work to investigate the outcome of the coalescence of a black hole with a neutron star , to find out to what extent the neutron star is tidally disrupted , and in particular to determine if an accretion structure does form around the black hole as a result of the encounter .
we also wish to explore the role of the mass ratio and the stiffness of the equation of state in the evolution of the system and in the emission of the gravitational radiation . for simplicity and to perform accurate comparisons with the work presented by @xcite , ( hereafter rs ) for double neutron star binaries , we have chosen to model the neutron star as a polytrope and explore different mass radius relationships by varying the stiffness through the adiabatic exponent @xmath1 . in this paper
we restrict ourselves to a stiff equation of state with @xmath1=3 .
we restrict our work to binaries which are tidally locked .
the construction of a completely self consistent initial condition is then entirely straightforward .
this assumption represents a major simplification in our calculations , and can be seen as an extreme case of angular momentum distribution in the system .
the opposite extreme would be a case where the binary components exhibit no rotation at all as viewed from an external , inertial frame of reference , and it poses a much greater problem regarding the construction of self consistent initial conditions in equilibrium . since complete tidal locking is not expected ( @xcite ) , the full range of initial configurations deserves to be explored .
the motivation for our work is presented in section [ motivation ] .
the numerical method we have used to carry out our simulations is described in section [ method ] .
calibration to previous work and our modeling of the black hole
neutron star system is presented in section [ calibration ] . in section [ results ]
we present our numerical results , and a discussion of these results follows in section [ discussion ] .
the orbital evolution of short period binary systems containing compact objects ( black holes or neutron stars ) will inevitably lead to coalescence due to orbital angular momentum loss to gravitational radiation .
such binary systems have been found ( psr 1913 + 16@xcite , psr 1534 + 12@xcite ) and the rate of inspiral matches the prediction of general relativity to high accuracy ( @xcite ) . the time to final merging in these binaries is less than the hubble time and so they will eventually coalesce . in the final stages of such a coalescence , a powerful burst of gravitational waves is expected , and as such these systems are primary candidate sources for detection by instruments such as ligo and virgo , expected to begin operation within a few years ( @xcite ) .
estimates for the event rate can be inferred from the statistics of the known hulse taylor type binaries , and a rate of @xmath2 to @xmath3 per galaxy per year can be expected ( @xcite ) , which would imply several coalescences per year could be observed out to a distance of 1 gpc .
these gravitational waves will certainly carry a copious amount of information about the emitting source . among other things
, they are expected to exhibit strong departure from point
mass behavior due to hydrodynamical effects and the finite size of the stars , as well as a dependence on the details of the equation of state for matter at nuclear densities .
a fully three dimensional hydrodynamical study is therefore essential to understand the gravitational radiation signal , and this information can be used to place constraints on the equation of state . for the typical masses and binary separations involved in very close binary systems ( a few solar masses and several tens of kilometers respectively )
, one should ideally solve the problem using general
relativistic formalism and include radiation processes .
this has not yet been realized , but some qualitative results can hopefully be obtained from a newtonian treatment
. it has been shown ( @xcite ) that at least in the newtonian case , hydrodynamic effects play an important role in the orbital evolution of the system . essentially , tidal interactions can make a close binary dynamically unstable for small enough separations on the order of a few stellar radii when the spin angular momentum becomes comparable to the orbital angular momentum .
this effect alone can produce orbital decay on a time scale comparable to that of angular momentum loss to gravitational waves .
the final stages of inspiral , and the coalescence of double neutron star binaries has been the object of numerous previous studies in a newtonian framework .
the gravitational radiation signal was studied by nakamura & oohara ( @xcite ) and oohara & nakamura ( @xcite , @xcite ) while davies _ et al . _
( @xcite ) considered the effect of different initial spin configurations on the outcome of the merger .
more recent simulations have calculated the neutrino emission as well as the gravitational waves from the coalescence ( ruffert _ et al . _
@xcite ) . the de stabilization effect of tidal forces in a newtonian framework has been investigated in great detail ( @xcite , rs ) using an analytical approach based on an energy variational method to examine equilibrium configurations and a three dimensional numerical treatment to study the hydrodynamical aspects of the coalescence , using throughout a polytropic equation of state .
there is yet another reason why these systems are the focus of intense study , since they have been suggested ( @xcite ) as a possible source for the production of gamma
ray bursts ( grbs ) .
it is well established that the grb distribution on the sky is isotropic and it is generally believed that their sources lie at cosmological distances ( @xcite ) , with redshifts on the order of unity ( @xcite ) .
the observed fluence of said bursts , @xmath4 to @xmath5 erg @xmath6 , and the rise times in their light curves imply that an energy release of at least @xmath7 erg takes place in a few milliseconds , in a region at most a few hundred kilometers across . to produce a grb
, this amount of energy , after being released in a primary event , must be transformed into @xmath8rays .
the relativistic blast wave model ( @xcite ) requires a relatively baryon free line of sight to the observer along which matter can be accelerated to speeds close to the speed of light .
the interaction of this outflow with the interstellar medium would then presumably lead to shock acceleration of electrons and emission of @xmath8rays through synchrotron radiation . in the double neutron star merger scenario
( @xcite ) the merger initially produces a burst of neutrinos and anti neutrinos .
however , newtonian calculations ( @xcite ) have shown that the energy release into neutrinos is insufficient to power the observed grbs .
general relativistic calculations carried out in a conformally flat metric ( @xcite ) have shown that in some cases each neutron star may collapse into a black hole many orbits prior to merging so that no blast wave will occur .
nevertheless , this could still produce a grb with a smooth time profile ( @xcite ) . in the black hole
neutron star merger scenario ( @xcite ) , it was expected that the neutron star will be disrupted by the black hole and a thick accretion torus will be formed . as in the case with two neutron stars ,
this could lead to the formation of a blast wave which would power the grb .
for the calculations presented in this paper , we have used the numerical technique known as smooth particle hydrodynamics ( sph ) .
this is essentially a lagrangian method , where forces are evaluated by interpolation over a grid of points co moving with the fluid , which can be considered as particles .
this method was originally developed by @xcite and @xcite as an alternative to eulerian hydrodynamics ( computations on a fixed grid ) .
the principal advantages of sph are that no assumptions need to be made _ a priori _ about the nature of the flow that will be studied and that no computational effort is wasted in modeling regions where matter is not present .
this is particularly advantageous when studying complicated three dimensional astrophysical flows .
an excellent review of the method has been given by @xcite .
sph has been tested successfully and applied to a variety of problems , such as the shock wave in a tube problem ( @xcite ) , redistribution of angular momentum in a thick accretion torus ( @xcite ) , static stellar structure ( @xcite ) , astrophysical jets ( @xcite ) , and hydrodynamics of close binary systems ( @xcite , @xcite , @xcite ) .
we have developed our own sph code ( @xcite ) and successfully tested it in one , two and three dimensions with several of these problems .
agreement with previous results has been excellent in every case .
a description of our code is given in appendix [ code ] .
given that the problem of merging black hole neutron star binaries is closely related to the study of coalescing double neutron star or white dwarf binaries , we have performed a rigourous calibration of our code to the results presented by rs for two polytropes with stiff equations of state ( @xmath1=3 ) and an initial mass ratio of unity in a tidally locked binary .
while our simulations had a lower resolution ( @xmath92000 particles per star vs. @xmath10 for rs ) , qualitative and quantitative agreement was excellent .
for the following , measure distance and mass in units of the radius and mass of the unperturbed ( spherical ) neutron star ( 13.4 km and 1.4@xmath11 respectively ) , except where noted , so that the units of time , density and velocity are : @xmath12 @xmath13 @xmath14 where @xmath15 and @xmath16 are the radius and mass of the unperturbed ( spherical ) neutron star .
the coalescence resulted in a massive central core containing about 82% of the total mass , surrounded by a massive halo and extended spiral arms . at the end of the calculation , the central object is not azimuthally symmetric and thus continues to emit gravitational waves , of amplitude @xmath17 .
the amplitudes and frequencies of the gravitational radiation waveforms were found to be in excellent agreement with rs ( see @xcite , @xcite ) . in table [ rslk ] we present a comparison of some of the more important parameters in the case of the coalescence of two neutron stars .
the first , second and third columns show the maximum and final amplitude in the gravitational radiation waveforms and the maximum gravitational wave luminosity respectively .
these values are given in geometrized units , where @xmath18=@xmath19=1 , and the peak luminosity @xmath20 is normalized to @xmath21=@xmath22=@xmath23 erg s@xmath24 .
the fourth and fifth columns display the central density of the resulting core and the mass contained in the halo surrounding it , as a fraction of the total mass in the system .
for the simulations of a black hole neutron star binary presented here , we have chosen to keep a stiff equation of state ( @xmath1=3 ) to compare our results to the case of the coalescence of two neutron stars . in future work , we will explore different values of the adiabatic index @xmath1 to model different mass radius relationships .
this is particularly important since future measurements of the gravitational radiation emitted during these events could serve to constrain the equation of state for dense matter . in our simulations ,
the black hole is modeled by a point mass , and produces a newtonian potential : @xmath25 the contribution from the black hole , to the force on particle _ i _ in the star is then simply given by : @xmath26 and symmetrically , @xmath27 is the contribution from particle _
i _ to the force on the black hole .
the horizon of the black hole is modeled by placing an absorbing boundary at a distance @xmath28=@xmath29 from the point mass . at every time step during the dynamical simulations , any particle that crosses this boundary
is absorbed by the black hole and removed from the simulation .
the mass , position and velocity of the black hole are adjusted so that total mass and linear momentum are conserved .
we thus disregard any spin angular momentum that might be gained by the black hole during the process .
this does not present any problems , since our calculation is newtonian throughout , and we make no attempt to model frame dragging .
our simulations begin with the construction of a spherical , unperturbed polytrope , which for simplicity we will refer to as the neutron star . to do this ,
we proceed essentially as rasio & shapiro ( 1992 ) .
a total of @xmath30 particles are placed on a cubic lattice with masses @xmath31=@xmath32 , where @xmath33 is the density calculated from the lane
emden solution to the equation of hydrostatic equilibrium with an equation of state @xmath34=@xmath35 and _
n _ is the number density of particles .
the smoothing length is assigned to each particle so that the number of overlapping neighbors per particle is @xmath36=64 .
this amounts to setting the smoothing length @xmath37 , where _ l _ is the lattice spacing .
there is considerable advantage in having variable particle masses , as this increases the spatial and density resolution near the edge of the star , where the density gradient is largest .
the radius of the unperturbed polytrope is 13.4 km , and its mass is @xmath38 . for the simulations presented in this paper , we have used @xmath39 for every case except two , where @xmath40 . to obtain tidally locked
equilibrium configurations for the binary system , the unperturbed polytrope and the black hole are placed in the co rotating keplerian frame of initial angular velocity @xmath41= @xmath42 , where _ r _ is the binary separation defined as the distance between the black hole and the center of mass of the neutron star .
a damping term linear in the velocity is introduced into the equations of motion for the sph particles to allow the star to respond to the presence of the tidal gravitational field .
we now have : @xmath43 @xmath44 where and are the gravitational and hydrodynamical forces on particle _ i _ ( of mass @xmath31 ) respectively .
the value of @xmath45 is chosen so that oscillations are critically damped , reaching an equilibrium as quickly as possible ; @xmath46 .
we neglect coriolis forces since we are interested in equilibrium configurations with no bulk motion in the co rotating frame .
the position of the black hole and the center of mass of the star are adjusted at every time step so that the binary separation remains at a desired value . in the same manner , @xmath41 is adjusted so that the total ( gravitational plus centrifugal ) force on the black hole in the co rotating frame , @xmath47 is zero , i.e. : @xmath48 where is the position vector of the black hole .
this procedure ensures that the configuration reaches equilibrium in a state of synchronization . for a given value of the binary separation ,
we allow the system to relax for a period of twenty time units , keeping the specific entropies of all particles constant , i.e. @xmath49=constant in @xmath34=@xmath35 . in all cases ,
our initial conditions satisfy the virial ratio to better than three parts in @xmath50 .
we have varied the mass ratio in the binary , @xmath51=@xmath52 , by changing the mass of the black hole only .
in this section we present the results of our simulations for different values of the mass ratio @xmath51 in the binary , beginning with high values of @xmath51 and proceeding in descending order .
the units used are as defined in equations ( [ eq : deftunit ] ) , ( [ eq : defrhounit ] ) and ( [ eq : defvunit ] ) except where noted . the highest value of the mass ratio is unity .
although the production of a binary system with such a low mass black hole is rather unlikely , we nevertheless wish to carry out a comparison with the case for two neutron stars , and use this as a starting point in our investigation . for each mass ratio we present the initial configurations that were constructed for tidally locked binaries , followed by a description of the dynamical runs . to investigate the dynamical evolution of the system for a given initial separation , we remove the damping term from the equations of motion and give every sph particle and the black hole the azimuthal velocity corresponding to the equilibrium value of @xmath41 in an inertial frame , with the origin at the center of mass of the system . each sph particle is assigned a specific thermal energy @xmath53=@xmath54 , and the equation of state is changed to that of an ideal gas where @xmath34=@xmath55 .
the specific thermal energy of each sph particle is then evolved individually throughout the simulation according to the first law of thermodynamics , taking into account the contribution from the viscous terms ( see appendix [ code ] ) .
as described above ( section [ init ] ) , equilibrium configurations were constructed for different values of the binary separation _
r_. we show in figure [ jvsrq1 ] a plot of the total angular momentum of the system , _ j _ , versus binary separation _ r_. the solid line is the result for two point masses in keplerian orbit .
the dotted line is the result of approximating the neutron star as a compressible tri axial ellipsoid ( @xcite ) , and one cross is plotted for each sph calculation at a fixed separation ( with @xmath30=8121 sph particles for the neutron star ) .
the presence of a minimum in _
j _ is a direct result of the assumption of tidal locking of the extended star . essentially , as the separation
is decreased , the spin component of angular momentum must grow ( just as the orbital component is decreasing ) , and eventually become comparable to the orbital component .
the result is that the total angular momentum _
j _ reaches a minimum and as the separation decreases further , it will increase .
such a turning point in the curve of total angular momentum as a function of separation marks the onset of an instability , which will lead to orbital decay ( @xcite ) . note the strong departure from point mass behavior , and the excellent agreement of the full sph calculations with the compressible tri axial ellipsoid treatment before the minimum in _ j _ is achieved .
close to the minimum , the ellipsoidal approximation breaks down and a full numerical treatment is necessary .
the cross with the smallest value of @xmath56 corresponds to roche lobe overflow ( @xmath56=@xmath57=2.78 ) .
our initial determination of the stability limit ( @xcite ) was higher than this value because our initial conditions did not correspond to proper tidal locking .
we show in figure [ initposq1 ] plots of particle positions projected onto the equatorial plane for different separations . as the separation
is decreased , the effect of the tidal potential becomes more apparent and the neutron star is distorted .
the configuration at @xmath56=3.0 can be approximated by a tri axial ellipsoid , while for those at @xmath56=2.8 and @xmath56=2.78 this is increasingly no longer the case .
we have used several of the configurations thus constructed to perform dynamical runs . in figure [ rvstq1 ]
we show the binary separation as a function of time during several of these calculations .
it is clear that the configurations with initial separations @xmath56=3.0 and @xmath56=2.9 are stable .
the separation exhibits oscillations of numerical origin on a period close to the orbital period ( respectively @xmath34=22.81 and @xmath34=21.6 in our units , see eq .
[ [ eq : deftunit ] ] ) . however , the configuration with @xmath56=2.78 and initial period @xmath34=20.09 ( corresponding to 37 km and 2.39 ms for @xmath15=13.4 km and @xmath16=@xmath38 ) is unstable , and the separation decays on an orbital timescale .
we find the dynamical stability limit to be @xmath58=2.78=@xmath57 .
the two different lines with initial separation @xmath56=2.78 correspond to two different resolutions ( the solid line used @xmath30=8121 particles and the dotted line used @xmath30=16944 particles ) . in the following , we will describe the results for our higher resolution run , with @xmath30=16944 . in figure [ rhobhq1 ]
we show density contours at various times during the simulation , which was run from @xmath59=0 to @xmath59=100 .
the left column shows contours in the orbital plane , and the right column shows contours in the meridional plane containing the black hole and the center of mass of the sph particles .
the black disk with radius @xmath28 represents the black hole .
mass transfer from the neutron star to the black hole begins almost immediately ( within one orbit ) through the inner lagrangian point .
the accretion stream gradually becomes thicker and begins to wrap around the black hole . by @xmath59=50
the star has become considerably elongated and a toroidal structure has begun to form around the black hole .
the accretion stream then essentially breaks , mass transfer stops and a stellar core remains in orbit around the black hole ( @xmath59=60 through @xmath59=70 ) .
the result of this coalescence then appears to be a a stable binary with a greatly altered mass ratio and separation ( @xmath60=0.19 and @[email protected] , corresponding to 60 km ) , consisting of a remnant core with mass @xmath61=0.307 ( corresponding to @xmath62 ) in orbit around a black hole with @xmath63=1.607 ( or @xmath64 ) .
the mass transfer event is very brief , as can be seen in figure [ mbhq1 ] , where we plot the mass of the black hole and the mass accretion rate onto the black hole as a function of time .
the peak accretion rate is @xmath65=0.0275 ( corresponding to @xmath66ms ) .
the black hole is still accreting matter towards the end of our simulation , but the decreased resolution ( recall that accretion entails a loss of particles in the simulation ) does not allow us to determine the distribution of matter accurately beyond @xmath59=100 ( @xmath911 ms ) .
what can be determined is that the toroidal structure that forms does _ not _ extend to form a halo that engulfs the black hole completely .
the region directly above the black hole remains devoid of matter down to the limit of our resolution of @xmath67 ( or @xmath68 ) within @xmath69 .
the accretion structure around the black hole was not previously observed ( @xcite ) because of the low resolution ( @xmath30=2176 particles ) of our initial runs .
the nature of the encounter is reflected in the gravitational radiation waveforms , presented in figure [ bhnsq1grav ] ( top ) .
the amplitude of the emitted waves initially rises from @xmath70=1.55 to @xmath71=1.75 , reflecting the decrease in separation .
the subsequent drop in amplitude and decrease in frequency are a result of the short episode of mass transfer , and the final waveforms reflect the fact that the binary has survived the encounter ( the new binary orbital period is @xmath72 in our units , or 4.5 ms ) .
thus the presence of a remnant core in orbit around the black hole would be immediately apparent from an observation of the waveforms emitted during such a coalescence .
we also show the gravitational wave luminosity in figure [ bhnsq1grav ] ( bottom ) . for comparison purposes ,
the luminosity for the coalescence of two identical neutron stars is plotted as well , from the results of section [ nsns ] .
the maximum luminosity and total energy radiated in gravitational waves are @xmath73=0.39 ( or @xmath74 erg s@xmath24 ) , @xmath75 ( or @xmath76 erg ) and @xmath77=0.15 ( or @xmath78 erg s@xmath24 ) , @xmath79 ( or @xmath80 erg ) for the double neutron star and black hole
neutron star case respectively .
although in both cases there is still a clear rise and decay in the luminosity curve , the peak is much broader ( @xmath81 ) than for two neutron stars ( @xmath82 ) .
the slower rise in luminosity and longer timescale of the peak in the black hole neutron star case are a direct consequence of having only one star being deformed and disrupted through tidal interactions . finally , we show in figure [ ptdecayq1 ] the binary separation as a function of time for the simulation we have just presented , together with a plot of the separation for a point
mass binary with the same mass ratio and initial separation decaying via the emission of gravitational waves in the quadrupole approximation .
this last curve is given by ( e.g. @xcite ) @xmath83 where @xmath84 and @xmath85 is the initial separation at @xmath59=0 .
the total mass is @xmath86=@xmath63+@xmath16 and the reduced mass is @xmath87=@xmath88 . clearly , hydrodynamical effects and the instability they induce drive the coalescence process on a time scale that is shorter than that purely due to gravitational radiation losses , and so we may safely neglect the effect of radiation reaction on the evolution of the system over the time period which we have modeled . in the calculation described above , we have integrated the thermal energy equation according to the first law of thermodynamics by setting @xmath89 where @xmath90 is calculated using the contribution from the artifical viscosity ( see equation [ eq : dudt ] in appendix [ code ] ) .
we will denote this as case i. to investigate the effect of radiative losses on the outcome of this configuration , we have performed an additional calculation for the extreme case in which all the energy produced by viscosity is lost by the system ( e.g. to neutrinos ) .
this will be denoted as case ii , and is achieved by simply removing the term associated with viscous heating from equation ( [ eq : thermoi ] ) , so that we now have @xmath91=@xmath92 .
density contours for case ii ( with @xmath30=16944 particles ) are shown in figure [ rhoii ] for @xmath59=60 and @xmath59=80 .
an accretion torus is again clearly visible , and comparing with the contours in figure [ rhobhq1 ] for case i , the ring is clearly more confined both in the orbital plane and along the @xmath93-axis .
this is clearly a consequence of a lack of pressure support due to the removal of thermal energy from the system . in figure [ heatloss ]
we show the rate of energy loss , @xmath94 due to viscous heating .
the peak rate is @xmath95 , corresponding to @xmath96 erg s@xmath24 and the total energy loss from @xmath59=0 to @xmath59=100 is @xmath97 erg .
the mass transfer rate and gravitational radiation waveforms are essentially the same as for case i. there is once more a baryon free axis along the rotation axis of the binary , only this time it is clear of matter within @xmath98 ( down to @xmath99 , equivalent to @xmath100 ) since the torus is more nearly confined to the orbital plane . for a mass ratio @xmath51=0.8 , the black hole mass is @xmath63=1.25 ( or 1.75@xmath11 ) .
hydrodynamical effects are still important in the construction of equilibrium configurations in this case , as can be seen in figure [ jvsrq08 ] , where we have plotted the equilibrium angular momentum values for a range of binary separations .
approximating the neutron star as a compressible tri axial ellipsoid gives results which are close to the full numerical calculations before the minimum in angular momentum is reached , but for @xmath101 this is no longer the case .
we identify the separation with minimum angular momentum at @xmath56=2.98 , and roche lobe overflow occurs at @xmath57=2.935 . for this value of the mass ratio @xmath30=8121 particles model the neutron star . as for the case with @xmath51=1 ,
several equilibrium configurations were used as initial conditions for dynamical runs .
figure [ rvstq08 ] shows the binary separation as a function of time for these calculations .
the initial configuration with separation @xmath56=2.935 becomes unstable on a dynamical time scale .
however , it is apparent that the development of the instability ( and hence the orbital decay ) proceeds at a slower rate than for a higher mass ratio .
the reason for this is that since the black hole is more massive , roche
lobe overflow occurs at a greater separation than previously observed ( for @xmath51=1 ) , and so the tidal effects are less pronounced on the neutron star .
density contours in the orbital plane for the binary with initial separation @xmath56=2.935 are shown in figure [ bhnsq08contours ] .
mass transfer proceeds through the formation of an accretion stream from the neutron star to the black hole , lasting from @xmath59@xmath945 to @xmath59@xmath975 ( about 3.5 ms ) , with a peak accretion rate of 0.017 ( 0.2@xmath11/ms ) .
a substantial amount of mass is stripped from the neutron star during the encounter ( see figure [ mbhq08 ] ) , and the final black hole mass is @xmath63=1.65 ( equivalent to 2.3@xmath11 ) .
the core of the neutron star , with a final mass @xmath61=0.595 ( corresponding to 0.83@xmath11 ) again survives the encounter .
however , the most striking difference between this case and the one presented previously ( @xmath51=1 ) is the absence of a toroidal accretion structure around the black hole .
all matter stripped from the neutron star is directly accreted by the black hole , and thus in this scenario as well the rotation axis above the black hole is devoid of matter . as a result of mass transfer the binary separation increases to @[email protected] and
thus the final configuration in this case is again a stable binary . as before , the hydrodynamical evolution of the binary is reflected in the gravitational radiation waveforms , presented in figure [ bhnsq08grav ] ( top ) .
there is an initial rise in amplitude in the waveforms , followed by a rapid decline and increase in the period , reflecting the episode of mass transfer and subsequent increase in binary separation .
the peak amplitude in this case is @xmath71=1.93 , and again since the binary survives the encounter , the final amplitude is not zero but @xmath102=1.25 .
the luminosity emitted in gravitational waves is shown in figure [ bhnsq08grav ] ( bottom ) .
the general shape of the curve resembles very much that for @xmath51=1 , as one can expect after studying the density contour plots for each case ( since the toroidal structure that was formed for @xmath51=1 is essentially azimuthally symmetric it does not contribute significantly to the emission of gravitational waves ) .
the curve presents again a broad peak , with a duration @xmath103@xmath930 ( 3.4 ms ) .
the peak luminosity is @xmath104=0.16 ( or @xmath105 erg s@xmath24 ) and the total energy radiated away in gravitational waves is @xmath106@xmath107@xmath108 ( or @xmath109 erg ) .
these values are higher than for the case with @xmath51=1 because we have normalized our results to the mass of neutron star , and lowering the mass ratio implies the system has a greater total mass . in figure [ ptdecayq08 ]
we compare ( as for @xmath51=1 in figure [ ptdecayq1 ] ) the orbital decay of this binary with that for a point
mass binary emitting gravitational waves in the quadrupole approximation .
it is apparent that hydrodynamical effects are less important than for @xmath51=1 in determining the initial orbital evolution , but are nevertheless substantial .
continuing the trend towards lower mass ratios , we chose a value of @xmath51=0.31 , giving @xmath63=3.22 ( corresponding to 4.51@xmath11 ) to study the evolution of a binary system where most of the mass is contained in the black hole .
we again used @xmath30=8121 particles to model the neutron star .
a qualitative difference appears in the plot of total angular momentum as a function of binary separation ( figure [ jvsrq031 ] ) . at a separation of @xmath56=3.76 ( or 50.5 km ) , roche
lobe overflow occurs , so that we can not construct a sequence with decreasing separation and constant mass ratio anymore , but this point is reached _ before _
the minimum in total angular momentum is attained , contrary to what occurred for @xmath51=1 and @xmath51=0.8 .
thus we expect all configurations down to the roche limit @xmath57=3.76 to be dynamically stable ( see the discussion for @xmath51=1 in section [ initq1 ] ) .
the binary separation for a dynamical run with an initial separation @xmath56=@xmath57=3.76 is shown in figure [ rvstq031 ] .
the initial orbital period is @xmath34=21.92 ( or 2.51 ms ) .
there is no longer orbital decay , but rather a slow increase in separation .
this is simply due to the fact that there is slow roche
lobe overflow from the neutron star onto the black hole , transferring @xmath110=0.0104 ( or 0.0146@xmath11 ) over @xmath94.5 orbital periods .
the neutron star is not disrupted at all during this process , as can be seen in figure [ rhobhq031 ] , where we plot density contours in the orbital plane over the course of the simulation , and in figure [ bhnsq031h ] , where we plot the gravitational radiation waveform @xmath111 for this case .
in fact , mass transfer is essentially conservative ( i.e. the total angular momentum is conserved ) , and this approximation alone can accurately account for the change in separation . to see this , consider the following .
conservative mass transfer in a point
mass binary satisfies @xmath112 , where _ a _ is the separation and @xmath87 is the reduced mass . since the total mass is conserved and @xmath113 for a keplerian binary , we find that @xmath114 . for the black hole
neutron star binary with initial mass ratio @xmath51=0.31 , we find that @xmath115^{2}=1.01453\end{aligned}\ ] ] and @xmath116 this is consistent with having conservative mass transfer despite the fact that the black hole
neutron star binary is not keplerian .
angular momentum is practically constant since there is very little accretion , and this represents the only loss of angular momentum ( via a spinning up of the black hole , which we do not model ) .
this last case is a clear indication of the limitations of the present method of studying the coalescence of a black hole with a neutron star . in the first place , since the orbit does not decay due to hydrodynamical effects , gravitational radiation reaction ( which would drive the binary mentioned above to coalesce in a time @xmath117=31.35 , correponding to 3.59 ms ) can not be ignored .
second , as is apparent in the density contour plots shown in figure [ rhobhq031 ] , the black hole has become almost as large as the neutron star itself ( the schwarschild radius is @xmath28=0.986 , or 13.2 km , at the start of the simulation ) , making the radius of the innermost circular stable orbit ( isco ) @xmath118=@xmath119=2.958 ( or 39.75 km ) .
this means a substantial fraction of the neutron star is already at a distance @xmath120 from the black hole . thus including relativistic effects in this case
becomes necessary in order to obtain meaningful results .
we present in table [ critical ] the binary separations @xmath57 and @xmath121 , corresponding to the roche limit and to the equilibrium configuration with minimum total angular momentum @xmath122 for each mass ratio . the lack of an entry for @xmath121 when @xmath51=0.31 reflects the fact that the orbits are stable when the separation is greater than that required for roche lobe overflow .
in table [ summary ] we present a summary for the three configurations we explored in this paper , as well as for the coalescence of two identical neutron stars .
the first column indicates if the coalescence involves two neutron stars or one neutron star and one black hole ; the second and third columns show the initial and final mass ratios respectively ; the fourth column indicates whether an accretion torus was formed ; the fifth , sixth and seventh columns show the maximum and final gravitational radiation amplitude and the maximum gravitational radiation luminosity respectively , all in geometrized units such that @xmath18=@xmath19=1 ; the eighth column shows the mass of the central object for the case of the double neutron star coalescence , and that of the remnant core in the case of the black hole
neutron star coalescence .
note that only in the simulation with a high mass ratio ( @xmath51=1 ) did an accretion structure appear around the black hole .
otherwise , all matter stripped from the neutron star was directly accreted by the black hole .
a striking result is that in every case , the neutron star is not completely disrupted , but a remnant core survives in orbit around the more massive black hole . due to the violent mass transfer episode ( for @xmath51=0.8 and @xmath51=1 ) ,
this core is transferred to a higher orbit on a time scale of one orbit . for now , we are unable to model the system on longer time scales than we have presented here , but presumably the emission of gravitational waves will make the orbit decay by decreasing the separation . clearly mass transfer will occur before the separation reaches zero , as soon as the neutron star overflows its roche lobe . for the equation of state we have considered ( with @xmath1=3 ) , the mass radius relationship is @xmath123 and one can estimate the size of the roche lobe as @xmath124=0.46@xmath125 ( @xcite ) . using the rate of angular momenutum loss calculated from the quadrupole approximation , given by ( @xcite ) @xmath126 we can find the separation at which mass transfer will begin anew , as well as the time it will take for the orbit to shrink to this size . applying this reasoning to the binary with initial @xmath51=1 after the initial mass transfer event , we find that a second episode of mass transfer will occur @xmath940 ms later , and for the case with initial @xmath51=0.8 , the delay will be @xmath94 ms .
thus the length of the coalescence process is extended from a few milliseconds to possibly several tens of milliseconds .
another important result is the fact that in every case , there is a baryon free line of sight to the black hole along the rotation axis of the binary throughout the simulation .
as stated in section [ dynq1 ] , the limit we infer for the amount of matter along this axis from our calculations is @xmath68 within about 5@xmath127 of the rotation axis .
it is set by the resolution in our calculation and thus represents an upper bound at this point .
for the calculations with a lower mass ratio ( @xmath51=0.8 and @xmath51=0.31 ) there was no visible accretion structure present around the black hole , and so the rotation axis was clear of matter as well .
these last two results may make the coalescence of a black hole with a neutron star a promising candidate source for the production of grbs ( @xcite ) .
the survival of the neutron star core suggests a different outcome than what was initially expected for the coalescence of a black hole with a neutron star ( @xcite ) , and identifies this process as the only one known capable of producing low mass neutron stars . the violent episode of mass
transfer strips the neutron star of so much matter that it may drive it below the minimum mass required for stability .
if this happens , then the core explodes ( @xcite ) in approximately 0.1 seconds after a slow expansion phase that may last 20 seconds ( @xcite ) .
the presence of the black hole would complicate the outcome of this event .
this work was supported in part through kbn grant 2p03d01311 and dgdpa unam .
the equations of motion in sph are essentially those of an n body problem , with each particle representing a fluid element .
thus the equations for particle _ i _ can be written as : @xmath128 @xmath129 where , and _
m_@xmath130 are the position , velocity and mass of particle _
i _ respectively , and and denote the gravitational and hydrodynamic forces . the density at the position of particle _
i _ is @xmath131 where @xmath132 is the symmetrized kernel for particles _ i _ and _ j_. with @xmath133 representing the smoothing length for particle _ i _ , we have @xmath134 to obtain a value of the pressure for any particle , we use the equation of state . for an ideal gas
this becomes @xmath136 where @xmath1 is the adiabatic index and @xmath53 is the thermal energy per unit mass for particle _
i_. includes the contribution from the pressure gradient and from artificial viscosity . in symmetrized form
, it can be written as @xmath137 where @xmath138 @xmath139 here = @xmath140 , = @xmath140 , @xmath141 and @xmath142 are constants , @xmath143 is the speed of sound at the position of particle _
i _ , and @xmath144=@xmath145 and @xmath146=@xmath147 . in the simulations presented in this paper
we have used @xmath141=1 , @xmath142=2 and @xmath148=@xmath149 .
the form of , and in particular the symmetrization of the pressure gradient term , ensures that forces between pairs of particles are symmetric , and so linear and angular momentum are automatically conserved .
the viscosity is introduced with a term linear in the velocity differences , which produces shear and bulk viscosity , and a quadratic term to handle high mach numbers in shocks , analogous to the von neumann richtmyer artificial viscosity used in finite difference methods ( @xcite ) . the value @xmath53 of the thermal energy per unit mass is evolved in time according to the the first law of thermodynamics , taking into account the contribution from the viscous terms .
the symmetrized form is : @xmath150 the thermal energy equation is integrated by essentially using a two step procedure ( @xcite ) in order to preserve accuracy .
this is necessary because the heating rate depends on the pressure , and via the equation of state , on the thermal energy itself .
individual smoothing lengths @xmath133 are adjusted so that every particle has an approximately constant number @xmath36 of hydrodynamical neighbors , i.e. , those for which @xmath151 .
this ensures that adequate spatial resolution is maintained and that the level of accuracy of the interpolations remains uniform throughout the fluid . if @xmath152 is the smoothing length for particle _
i _ at step _
n _ , the value at step @xmath153 is found according to ( @xcite ) : @xmath154 , \label{eq : dhdt}\end{aligned}\ ] ] where @xmath155 is the number of hydrodynamical neighbors at step _
n_. we have found that a value of @xmath36=64 yields adequate sampling of the fluid without requiring excessive computation time .
gravitational force calculations constitute the most time consuming part of an @xmath30body code .
a direct calculation requires @xmath156 operations , and this becomes prohibitive as the number of particles is increased above a few hundred . in order to make this process more efficient
we have incorporated a binary tree structure ( @xcite ) into our code to obtain gravitational forces .
this makes the number of required operations of @xmath157 ) , thus allowing for simulations with several thousand particles .
the tree structure is built by pairing two particles into a new pseudo particle , called a node , if they are mutual nearest neighbors , and proceeding with this pairing over all particles and newly created nodes until there is only one node , at the center of mass of the particles being considered .
this hierarchical association makes up the binary tree which we use to compute the forces ( for a detailed description see @xcite ) .
the force on any particle can then be obtained by adding up contributions from particles close enough to be considered individually and by more distant aggregations , which can be described in terms of their multipolar expansion up to the quadrupole term . to decide if an aggregation of particles must be resolved to obtain the force on particle _ i _
, we consider the ratio @xmath158 , where _ d _ is the distance from particle _
i _ to the aggregate and _ s _ is its size . if @xmath159 , where @xmath160 is a dimensionless tolerance parameter , the aggregate is resolved further .
if the converse is true , the contribution to the force is obtained from the quadrupole expansion .
we adopt @xmath160=0.8 , which provides an accuracy for the force of better than one part in @xmath161 .
however , the aggregate is resolved further if it contains any individual particles which may be hydrodynamical neighbors for particle _ i_.
the subroutine corresponding to this approach simultaneously provides and the hydrodynamical neighbors for every particle that are used to obtain @xmath162 and @xmath163 in equations ( [ eq : defhydrof ] ) and ( [ eq : dudt ] ) .
our calculations of the gravitational forces are completely newtonian .
we use a leapfrog algorithm accurate to second order to integrate the equations of motion .
the time step is adaptive , and is taken to satisfy a combination of the courant criterion for stability and a further restriction on the maximum change allowed in velocity for any particle during one time step to conserve accuracy .
we follow @xcite and set the time step to be @xmath164=@xmath165 with @xmath166 @xmath167 one of the main objectives of the present work is to investigate the gravitational waves emitted by coalescing binaries . in our newtonian simulations
we calculate the waveforms of the emitted waves and the total luminosity in the quadrupole approximation . the amplitude of the retarded wave at time _
t _ and a distance @xmath168 away from the source is given by ( @xcite ) @xmath169 @xmath170 is the transverse traceless part of the reduced quadrupole moment , @xmath171
@xmath172 @xmath173 the summation is over the sph particles and the superscripts indicate the cartesian components .
we identify the gravitational acceleration on particle _ i _ as .
the term @xmath179 only includes the contribution from the sph particles contained in the neutron star .
we add the terms arising from the presence of the black hole as a point mass contribution : @xmath180 to obtain the total value of @xmath181 as @xmath182 99 benz , w. , bowers , r.l .
, cameron , a.g.w . , press , w.h .
1990 , astrophysical journal , 348 , 647 .
bildsten , l. , cutler , c. 1992 , astrophysical journal , 400 , 175 .
blinnikov , s.i . , novikov , i.d . ,
perevodchikova , t.v . ,
polnarev , a.g .
1984 , soviet astronomy letters , 10(3 ) , 177 .
coleman , c.s . ,
bicknell , g.v .
1985 , monthly notices of the royal astronomical society , 214 , 337 .
colpi , m. , shapiro , s.l . , teukolsky , s.a .
1991 , astrophysical journal , 369 , 422 .
davies , m.b . , benz , w. , piran , t. , thielemann , f.k .
1994 , astrophysical journal , 431 , 742 .
eichler , d. , livio , m. , piran , t. , schramm , d.n .
1988 , nature , 340 , 126 .
finn , l.s .
1989 , in frontiers of numerical relativity , ed .
evans , l.s .
finn & d.w .
hobill ( cambridge : cambridge univ .
press ) , 126 fishman , g.j .
, meegan , c.a .
1995 , annual review of astronomy and astrophysics , 33 , 415 .
gingold , r.a .
, monaghan j.j . 1977 , monthly notices of the royal astronomical society , 181 , 375 .
hernquist , l. , katz , n. 1989 , astrophysical journal supplement , 70 , 419 .
hulse , r.a .
, taylor , j.h .
, 1975 , astrophysical journal , 195 , l51 janka , h.t . ,
ruffert , m. 1996 , astronomy and astrophysics , 307 , l33 .
kluniak , w. , lee , w.h .
, astrophysical journal letters , 494 , l53 , 1998 .
lai , d. , rasio , f.a . ,
shapiro , s.l .
1993a , astrophysical journal letters , 406 , l63 .
lai , d. , rasio , f.a . ,
shapiro , s.l . 1993b , astrophysical journal supplement , 88 , 205 .
lattimer , j.m . , schramm , d.n .
1976 , astrophysical journal , 210 , 549 .
lee , w.h .
1998 , phd thesis , university of wisconsin madison .
lee , w.h . ,
kluniak , w. 1995 , acta astronomica , 45 , 705 .
lucy , l. 1977 , astronomical journal , 82 , 1013 metzger , m.r .
_ 1997 , nature 387 , 878 .
mszros , p. , rees , m.j .
1992 , monthly notices of the royal astronomical society , 257 , 29p .
mszros , p. , rees , m.j .
1993 , astrophysical journal , 405 , 278 .
misner , c.w . ,
thorne , k.s .
, wheeler , j.a .
1970 , gravitation ( ny : freeman ) .
monaghan , j.j .
1989 , j. comput .
monaghan , j.j .
1992 , annual review of astronomy and astrophysics , 30 , 543 .
monaghan j.j . , gingold , r.a .
1983 , journal of computational physics , 52 , 375 .
monaghan j.j . ,
lattanzio , j.c .
1985 , astronomy and astrophysics , 149 , 135 .
nakamura , t. , oohara , k. 1989 , progress of theoretical physics , 82 , 1066 .
narayan , r. , paczyski , b. , piran , t. 1992 , astrophysical journal , 395 , l83 .
narayan , r. , piran , t. , shemi , a. 1991 , astrophysical journal letters , 379 , l17 .
oohara , k. , nakamura , t. 1989 , progress of theoretical physics , 82 , 535 .
oohara , k. , nakamura , t. 1990 , progress of theoretical physics , 83 , 906 .
paczyski , b. 1967 , acta astronomica , 17 , 287 .
paczyski , b. 1986 , astrophys .
j. , 308 , l43 .
paczyski , b. 1991 , acta astronomica , 41 , 257 .
rasio , f. , shapiro , s.l .
1992 , astrophysical journal , 401 , 226 .
rasio , f. , shapiro , s.l .
1994 , astrophysical journal , 432 , 242 .
richtmyer , r.d . ,
morton , k.w . , 1967 , _ difference methods for initial value problems _
( new york , wiley ) ruffert , m. , janka , h .- th . ,
schfer , g. 1996 , astronomy and astrophysics , 311 , 532 .
ruffert , m. , janka , h .- th . , takahashi , k. , schfer , g. 1997a , astronomy and astrophysics , 319 , 122 .
ruffert , m. , rampp , m. , janka , h .- th . 1997b , astronomy and astrophysics , 321 , 991 .
shapiro , s.l . , teukolsky , s.a .
1983 , _ black holes , white dwarfs and neutron stars _ , wiley interscience .
sumiyoshi , k. , yamada , s. , suzuki , h. , hilldebrandt , w. , astronomy and astrophysics , 334 , 159 , 1998 .
taylor , j.h . ,
weisberg , j.m .
1982 , astrophysical journal , 253 , 908 .
taylor , j.h . ,
wolszczan , a. , damour , t. , weisberg , j.m . , 1992 ,
nature , 355 , 132 .
thorne , k.s .
1995 , in _ proceedings of the snowmass 95 summer study on particle and nuclear astrophysics and cosmology _ , eds .
e.w . kolb and r. peccei ( world scientific , singapore ) .
wheeler , j.a .
1971 , pontificae acad .
scripta varia , 35 , 539 .
wilson , j.r . ,
salmonson , j.d . , mathews , g.j .
1997 , in gamma ray bursts , c.a .
meegan , r.d .
preece , t.m .
koshut , eds . , ( aip : new york ) , 788 .
wilson , j.r . ,
mathews , g.j .
, maronetti , p. 1996
, physical review d , 54 , 1317 .
wolszczan , a. , 1991 , nature , 350 , 688 .
urek , w.h . ,
& benz , w. 1986 , astrophysical journal , 308 , 123 .
ns ns & 1.000 & 1.000 & no & 2.20 & 0.20 & 0.390 & 0.82 bh
ns & 1.000 & 0.190 & yes & 1.75 & 0.55 & 0.145 & 0.3070 bh
ns & 0.800 & 0.360 & no & 1.93 & 1.25 & 0.160 & 0.5950 bh
ns & 0.310 & 0.306 & no & 3.59 & 3.48 & 0.425 & 0.9896 | we present a detailed study of the hydrodynamical interactions in a newtonian black hole
neutron star binary during the last stages of inspiral .
we consider close binaries which are tidally locked , use a stiff equation of state ( with an adiabatic index @xmath0 ) throughout , and explore the effect of different initial mass ratios on the evolution of the system .
we calculate the gravitational radiation signal in the quadrupole approximation .
our calculations are carried out using a smooth particle hydrodynamics ( sph ) code .
subject headings : binaries : close gamma rays : bursts hydrodynamics stars : neutron |
for the past few decades magnetic nanoparticles have been attracting the attention of scientists from diverse disciplines from the standpoints of both fundamental understanding and useful applications @xcite .
magnetic nanoparticles have been under scrutiny since the days of nel @xcite and brown @xcite who developed the theory of magnetization relaxation for noninteracting single domain particles .
effect of inter - particle interaction on magnetic properties of several nanoparticle systems are available in the literature @xcite . in 1961
nel suggested that small particles of an antiferromagnetic material should exhibit superparamagnetism and weak ferromagnetism @xcite .
antiferromagnetic nanoparticles are known to show more interesting behavior compared to ferro and ferrimagnetic nanoparticles one of which is that the magnetic moment of tiny antiferromagnetic particles increases with increasing temperature @xcite .
this is quite unlike what is seen in ferro and ferrimagnetic particles .
real magnetic nanoparticles have disordered arrangement , distribution in size and random orientation of magnetization , making their behavior very complex and challenging to understand .
bulk nio has a rhombohedral structure and is antiferromagnetic below 523 k whereas it has a cubic structure and is paramagnetic above that temperature @xcite .
a bulk antiferromagnet has zero net magnetic moment in zero applied field .
if the surface to volume ratio , which varies as the reciprocal of the particle size , for an antiferromagnetic particle becomes sufficiently large then the particle can have a nonzero net magnetic moment because of uncompensated spins at the surface . according to nel
the moment due to uncompensated spins would be parallel to the axis of antiferromagnetic alignment .
if we consider a collection of such particles of varying sizes and shapes we will see that there will be a distribution of magnetic moments of different sizes , oriented randomly and interacting with each other magnetically .
thus the magnetic properties of a collection of such particles would be very different from that of the corresponding bulk material .
richardson and milligan @xcite were the first to report magnetic susceptibility measurements as a function of temperature for nio nanoparticles of different sizes .
they measured the magnetic susceptibility at a field of 3500 g and found a peak in the susceptibility much below the nel temperature and the peak temperature was found to decrease with decreasing particle size .
they attributed the observed behavior to the modified magnetic environment of the ni@xmath0 ions at the surface of the particles and the enhanced surface to volume ratio .
in this paper we present a detailed study on nio nanoparticles based mainly on magnetic measurements .
in the course of this work we will come across phenomena which indicate that a collection of nio nanoparticles behaves as a spin glass or a superparamagnet . in each of those cases we shall try to critically examine the data in the light of current understanding of the physics involved .
nio nanoparticles are prepared by a sol - gel method by reacting in aqueous solution , at room temperature , nickel nitrate and sodium hydroxide at ph = 12 as described elsewhere @xcite . in this work we used nickel ( ii ) nitrate hexahydrate ( 99.999% ) , sodium hydroxide pellets ( 99.99% ) , both from aldrich , and triple distilled water to make nickel hydroxide .
the samples of nickel oxide nanoparticles are prepared by heating the nickel hydroxide at a few selected temperatures for 3 hours in flowing helium gas ( 99.995% ) .
all the magnetic measurements are done with a squid magnetometer ( quantum design , mpms xl ) .
the average crystallite size is calculated by x - ray diffraction line broadening using the scherrer formula @xcite @xmath1 where @xmath2 is the wavelength of the x - ray ( 1.542 ) , 2@xmath3 is the bragg angle , @xmath4 is the full width at half maximum ( fwhm ) of a peak in radians and @xmath5 is the fwhm of the same peak of a standard sample .
we used specpure grade nio powder from johnson matthey & co. ltd .
( uk ) as the standard
. peaks ( 111 ) , ( 200 ) and ( 220 ) are used to calculate the average crystallite size .
the use of @xmath6 instead of @xmath4 in the scherrer formula takes care of instrumental broadening .
the crystallite sizes of nio samples prepared by heating ni(oh)@xmath7 at 250 , 300 , 350 and 700 @xmath8c turn out to be 5.1 , 6.2 , 8.5 and @xmath9 nm respectively and these numbers will be referred to as the average crystallite size in this paper .
transmission electron micrograph of nio sample prepared by heating ni(oh)@xmath7 at 250 @xmath8c gives mean particle size to be 5.6 nm with a standard deviation of 1.3 nm . we note that the mean particle size determined by transmission electron micrograph is very close to the average crystallite size determined by x - ray diffraction using the scherrer formula ( 5.1 nm ) which implies that on an average each nio nanoparticle is a crystallite ( tiny single crystal )
. more details of sample characterization like x - ray diffraction pattern , transmission electron micrograph , selected area electron diffraction pattern and particle size distribution are available elsewhere @xcite . the temperature dependence of _ ac _ susceptibility is measured at several frequencies ranging from 1 hz to 1 khz .
the data are taken as described below .
the sample is first cooled from room temperature to 10 k in a zero magnetic field .
then a probing _ ac _ magnetic field of 1.0 g amplitude is applied to measure the susceptibility as the temperature is slowly raised in short steps to 300 k. figure [ fig : ac - suscept - frequency ] shows the real part , @xmath10 , of the @xmath11 susceptibility of the 5.1 nm sample ; in the inset we show the imaginary part , @xmath12 .
we note that all the curves have a peak at some temperature ; as the frequency is raised the value of @xmath10 decreases and the temperature of the peak increases .
the behavior observed here is characteristic of both superparamagnets and spin glasses .
the cusp in the @xmath11 susceptibility seen in canonical spin glasses is not observed here .
the inset shows that below the peak temperature , @xmath12 does not depend on frequency . such frequency
independent @xmath12 has been observed in other nanoparticle systems as well@xcite .
a quantitative measure of the peak temperature shift with frequency is the relative shift in peak temperature , @xmath13 , per decade of frequency .
for the 5.1 nm sample this quantity turns out to be 0.018 . for many canonical spin glasses it lies between 0.0045 and 0.06 whereas for a known superparamagnet a-(ho@xmath14o@xmath15)(b@xmath14o@xmath15 ) a value of 0.28
has been reported @xcite .
ferritin , an antiferromagnetic nanoparticle , which has been found to be superparamagnetic , has a value of @xmath16 for this quantity@xcite .
we note that our value of 0.018 falls in the spin glass range .
small particles of antiferromagnetic materials are expected to be superparamagnetic just as small particles of ferro or ferrimagnetic materials .
that is , each particle would behave as a single giant moment and their magnetic susceptibility as a function of temperature would be curie - like at sufficiently high temperature .
the giant moment arises because of uncompensated spins at the surface of the particle .
an antiferromagnetic particle with uni - axial anisotropy has two low energy states , separated by an energy barrier @xmath17 , corresponding to the parallel or anti parallel alignment of the magnetization of the particle with the axis of antiferromagnetic alignment@xcite .
the susceptibility has a maximum at a certain temperature called the blocking temperature , below which the probability of thermally assisted transitions between the two low energy states decreases progressively towards zero .
the blocking temperature , @xmath18 , should increase with increasing particle size because the energy barrier separating the low energy states is proportional to the volume of the particle .
in fact @xmath19 where @xmath20 is the volume of a single particle , @xmath21 is a constant and @xmath22 is the field of measurement@xcite . the dynamics of superparamagnets
is described by arrhenius law @xmath23 where @xmath24 is the energy barrier for flipping the magnetic moment of the particle and @xmath25 , an attempt frequency , usually has a value in the range @xmath26 to @xmath27 hz and is related to the intra - potential - well dynamics@xcite .
if we fit the data shown in figure [ fig : ac - suscept - frequency ] to the arrhenius equation , where @xmath28 corresponds to the peak in the susceptibility curve taken at frequency @xmath29 , we get @xmath30 k and @xmath31 hz ; these numbers are too large and rather unphysical . unreasonable numbers like these are usually seen in spin glass systems@xcite while superparamagnets tend to give more reasonable numbers such as @xmath32 k and @xmath33 hz for ferritin@xcite . here
we have another indication that the nio particles are not behaving like a superparamagnetic system in the region of temperature where they have a susceptibility peak .
this fact and the previously noted peak temperature shift with frequency suggest that the maximum in the _ ac _ susceptibility curve might have its origin in spin glass like freezing rather than superparamagnetic blocking .
field of 1 g amplitude and 10 hz frequency .
the dotted line drawn to pass through the peaks shows that the peak temperature increases with decreasing particle size . ] in figure [ fig : ac - suscept - particle - size ] we compare @xmath10 for 5.1 , 6.2 , 8.5 and @xmath34 100 nm samples at 10 hz .
it is clear that the temperature of the peak decreases with increasing particle size .
this behavior is quite contrary to what is expected in superparamagnetic systems as described by equation [ super - para - blocking ] where @xmath35 volume of particle .
we see that just as the frequency dependence of the peak temperature was unlike that of a superparamagnet , its size or volume dependence is also unlike that of a superparamagnet . in this paper
we make an attempt to understand this non - intuitive behavior of nio nanoparticles .
we also note from figure [ fig : ac - suscept - particle - size ] that the magnitude of the susceptibility increases with decreasing particle size .
this is as one would expect since the relative number of surface spins and hence the number of uncompensated surface spins would increase with decreasing particle size .
our data is in qualitative agreement with the claim that in nio nanoparticles susceptibility at room temperature varies as @xmath36 where @xmath37 is the particle diameter@xcite .
owing to the too few number of data points in our data set we are not able to make any quantitative claim .
are plotted against @xmath38 in the inset .
the solid line shows a linear fit to the data with coefficient of determination @xmath39 . ]
we measured the @xmath11 susceptibility at various bias fields @xmath22 with an @xmath11 field of 1 g amplitude and 10 hz frequency .
this is shown in figure [ fig : ac - suscept - field ] . as @xmath22 increases , the peak temperature and peak value decreases .
we note that the peak temperature , @xmath40 , decreases as , @xmath41 , as shown in the inset of figure [ fig : ac - suscept - field ] .
this dependence corresponds to the so called de almeida - thouless ( at ) line @xcite given by : @xmath42 the extrapolation of the at line back to @xmath43 gives the spin glass transition temperature @xmath40 which , in this case , turns out to be about 167 k. compliance of the data with the at line is considered to be a strong evidence for the existence of a spin glass phase and it has been observed in different kinds of spin glasses @xcite .
recently , it has been used as an evidence for spin glass phase in @xmath44 nanoparticles @xcite as well as in thin films @xcite .
figure [ fig : dc - mag - temp ] shows the temperature dependence of zero field cooled ( zfc ) and field cooled ( fc ) magnetization in a 100 g @xmath45 field for the 5.1 nm particles .
generally , the zfc susceptibility shows a peak for both superparamagnets and spin glasses .
in contrast , it is usually seen that the temperature dependence of the fc susceptibility becomes saturated below the peak temperature ( freezing temperature @xmath40 ) for spin glasses and continues to increase below that temperature ( blocking temperature @xmath46 ) for superparamagnets @xcite .
nevertheless , we note that , glassy behavior in magnetic nanoparticles has also been claimed in which the fc susceptibility continues to increase with decreasing temperature @xcite . in our case , below the peak temperature , the susceptibility continues to increase but with a tendency towards saturation .
this behavior is rather ambiguous .
figure [ fig : dc - mag - particle - size ] shows our data on the temperature dependence of zfc magnetization as a function of particle size in an applied field of 100 g. clearly the peak temperature decreases with increasing particle size , as has been seen in _ ac _ susceptibility measurements in section [ subsec : ac - suscept - particle - size ] earlier .
again , we notice the non - superparamagnetic behavior . a similar observation has been reported recently for mno nanoparticles where the temperature of the peak in @xmath45 susceptibility in a zero field cooled measurement decreases with increasing particle size @xcite .
the behavior we see here at 100 g field does not agree with the observations reported by richardson and milligan@xcite on nio where they found that at 3500 g the peak temperature in _ dc _ susceptibility increases with increasing particle size , as already mentioned in the introduction .
we did @xmath45 susceptibility measurement at 3500 g and found that we are able to reproduce@xcite the data of richardson and milligan .
we shall see later in section [ further discussion ] that these two conflicting observations are actually consistent with each other . in figure
[ fig : dc - mag - hysteresis ] we show the m - h curves obtained from _
magnetization measurements at 10 k and 300 k for the 5.1 nm particles .
hysteresis loop is seen in the 10 k data .
the hysteresis could be an indication of the weak ferromagnetism alluded to by nel in his musings on antiferromagnetic particles @xcite ; or it could be because the magnetization is time dependent and is relaxing too slowly for the experimental time scale of a few hours .
we shall soon see that our data support the latter proposition . at 300 k
there is no coercive force and no hysteresis .
we carried out a few more @xmath47 vs. @xmath22 measurements at different temperatures up to 350 k and found that the magnetization is not a function of @xmath48 as one would expect for superparamagnetic systems .
similar behavior has been reported by others also on nio nanoparticles @xcite . in figure
[ fig : dc - mag - fc - zfc ] we show the zfc and fc magnetization curves for the 5.1 nm sample at various applied fields . as the measuring field increases we find that both the bifurcation temperature of zfc and fc curves and the peak temperature of zfc curve shift to low temperature .
in fact , at 20 kg the zfc curve does not peak even on going down to 10 k. this behavior is along expected lines since we would expect in a sufficiently high external magnetic field the spin glass freezing will either take place at a lower temperature or not at all as a sufficiently strong external magnetic field can break the spin glass phase . [ hbt ]
thermoremanent magnetization ( trm ) is the time dependent remanent magnetization obtained when a sample is cooled from a temperature well above @xmath40 to a temperature below @xmath40 in an applied field @xmath22 and subsequently the field is removed . in this work , to measure the trm , we cool the sample from 300 k to the temperature of interest in a 100 g magnetic field .
a spin glass system is not in thermal equilibrium and slowly relaxes to lower energy states .
many models have been proposed to describe the time dependent relaxation of magnetization in spin glasses .
a single energy barrier should give an exponentially relaxing trm , with time constant @xmath49 , ie .
, @xmath50 . in a real system
there would be a range of energy barriers and hence there would be a distribution of relaxation times .
one of the popular forms incorporating a distribution of relaxation times is the stretched exponential function : @xcite @xmath51 where @xmath49 is a characteristic relaxation time and @xmath52 and @xmath53 are fit parameters . by assuming that the energy barriers are uniformly distributed from zero to some maximum energy
, it has been shown that the magnetization decays logarithmically @xcite as @xmath54 where @xmath52 and @xmath55 are fit parameters .
it should be noted that @xmath56 is the magnetization at @xmath57 unit and hence it depends on the unit of time used , while @xmath55 itself does not have any such dependence .
a power law form has also been used to describe the decay of magnetization : @xcite @xmath58 where the exponent @xmath53 should increase with increasing temperature . in this case also , as in the previous equation , @xmath56 depends on the unit of time .
a statistical measure of the goodness of a fit is @xmath59 , defined as @xcite @xmath60}^{2}}{\sigma_{i}^{2}}\ ] ] where n is the total number of data points and @xmath61 is the standard deviation of the @xmath62 data point . for a good fit
@xmath59 should have a value close to n. we shall also be reporting the coefficient of determination @xmath63 for our fits ; the closer @xmath63 is to unity the better the fit .
we fitted our data to the equations ( [ eqn : stretched - exp ] ) , ( [ eqn : log - decay ] ) and ( [ eqn : power - law ] ) . to compare the fits for the different forms we present the values of @xmath59 and @xmath63 for a representative data set , the 100 k trm data for 5.1 nm particles , in table [ table - comparison - of - forms ] .
we note that the stretched exponential gives the best fit with the lowest @xmath59 and largest @xmath63 .
this is followed by the logarithmic form with the power law form bringing up the rear . to check whether the fit parameters truly characterize the data we decided to run the fits on subsets of the original data covering different time spans such as ( 150 s , 1000 s ) , ( 150 s , 2000 s ) , ( 150 s , 3000 s ) and so on .
the fit parameters are found to be strongly dependent on the time span of the data for the stretched exponential case but are found to be almost independent of the same for logarithmic and power law cases .
consequently the stretched exponential fit has to be rejected on the grounds that the fit parameters in that case are simply artifacts of the subset of the data used , even though it has the best @xmath59 and @xmath63 values .
.values of @xmath59 and the coefficient of determination @xmath64 obtained by fitting 100 k trm data to various expressions for 5.1 nm sample .
total number of data points , @xmath65 , is 749 . [
cols="^,^,^",options="header " , ] in sections [ subsec : ac - suscept - particle - size ] and [ subsec : dc - mag - particle - size ] we saw that the peak temperature of the low field susceptibility as a function of temperature decreases with increasing particle size .
we take this to mean that the behavior of the system is not superparamagnetic and the peaks the in low field susceptibility vs. temperature curves are not because of superparamagnetic blocking of particle magnetic moments . from sections [ subsec : t&fdepend - ac - chi ] ,
[ subsec : field - depend - ac - chi ] and [ subsec : time - depend - trm ] it is clear that the system is showing spin glass behavior . now the question is , what is the mechanism behind the spin glass freezing ? there are reports in the literature where dipolar interaction between particles has been proposed as the reason for the freezing of particle magnetic moments @xcite . now let us examine whether there is such a possibility in the case of nio .
we estimate that for the 5.1 nm particles there would be an average uncompensated moment of the order of @xmath66 @xcite .
the maximum dipolar interaction energy between two such particles touching each other would be @xmath67 erg which corresponds to about 0.1 k on temperature scale .
this means that if dipolar interaction were causing the freezing it would occur at about 0.1 k which is much lower than the observed freezing temperature of about 167 k ( from @xmath11 susceptibility ) .
thus we rule out the possibility that dipolar interaction among particles is causing the peaks in low field susceptibility as a function of temperature . ) of zfc and fc magnetization curves
is considered to be the starting point of the freezing process .
the peak temperature ( @xmath68 ) in the zfc magnetization curve is generally considered as the freezing temperature .
this figure shows plots of @xmath69 and @xmath68 vs. @xmath38 .
the solid lines show linear fits to the data with @xmath63 of 0.9984 and 0.9954 respectively .
the upper and lower curves have zero field intercepts of 297 k and 153 k respectively . ] in canonical spin glasses the spin glass phase is very sensitive to external fields and the application of a field of a few hundred gauss breaks it and makes it disappear .
but here it is not the case and the spin glass phase survives up to the highest field used to measure zfc and fc magnetization ( up to 20 kg in our study ; see figure [ fig : dc - mag - fc - zfc ] ) .
the existence of spin glass phase up to such high fields is also evident from the at lines shown in figure [ fig : high - field - at - line ] .
clearly the peak temperature of the susceptibility curve decreases with increasing applied field following the at line .
we also find that the rate of change with field of the peak temperature of the susceptibility curve decreases with increasing particle size ( not shown here ) . to paraphrase , for smaller particles the @xmath68 changes more with field than for the larger particles .
it is this which gives rise to and resolves the conflicting observations noted at the end of section [ subsec : dc - mag - particle - size ] .
kodama et al .
@xcite and martnez et al .
@xcite have argued that freezing of surface spins can lead to spin glass phase which survives up to such high fields .
kodama et al .
@xcite have proposed a model for ferrimagnetic nickel ferrite nanoparticles where the spins at the surface of a particle are disordered , leading to frustration and spin glass like freezing .
a similar model has also been proposed for antiferromagnetic ferrihydrite nanoparticles @xcite .
we feel that such a model may be applicable in the case of nio nanoparticles as well .
we will first have to see whether we can propose any mechanism for surface spin disorder in our case .
the exchange interaction between two neighboring ni@xmath70 ions is mediated by an oxygen ion ( superexchange ) , and if an oxygen ion is missing from the surface , the exchange bond would be broken and the interaction energy would be reduced .
also , the average coordination number for ni@xmath70 ions at the surface will be less than that in the bulk and this can result in a distribution of exchange energies for the surface spins .
moreover , the superexchange is sensitive to bond angles and bond lengths , which are likely to be different at the surface compared to that in the bulk .
the reasons mentioned above may be sufficient to give rise to surface spin disorder and frustration leading to a spin glass phase .
the fraction of atoms lying on the surface of the particles increases with decreasing particle size .
this may lead to increased surface spin disorder as the particle size decreases and could possibly account for the increasing freezing temperature with decreasing particle size .
in this paper we reported a detailed study on magnetic properties of sol - gel prepared nio nanoparticles .
the behavior of the system is not superparamagnetic as was expected .
in fact it shows spin glass behavior which we attribute to surface spin disorder .
j. o. anderson , c. djurberg , t. jonsson , p. svedlindh , and p. nordblad , phys .
b * 56 * , 13983 ( 1997 ) ; c. djurberg , p. svedlindh , p. nordblad , m. f. hasen , f. bdker , and s. mrup , phys . rev . lett .
* 79 * , 5154 ( 1997 ) ; j. l. dormann , d. fiorani , r. cherkaoui , e. tronc , f. lucari , f. dorazio , l. spinu , m. nogus , h. kachkachi , and j. p. jolivet , j. magn .
magn . mater . *
203 * , 23 ( 1999 ) ; j. l. dormann , d. fiorani , r. cherkaoui , l. spinu , f. lucari , f. dorazio , m. nogus , e. tronc , j. p. jolivet , and a. garcia , nanostructured materials * 12 * , 757 ( 1999 ) . s. mrup , m. b. madsen , j. franck , j. villadsen , and c. j. w. koch , j. magn .
* 40 * , 163 ( 1983 ) ; i. tamura and m. hayashi , j. magn .
magn . mater . *
72 * , 285 ( 1988 ) ; d. h. reich , b. ellman , j. yang , t. f. rosenbaum , g. aeppli , and d. p. belanger , phys .
b * 42 * , 4631 ( 1990 ) .
it has been reported that 15 nm particles of nio prepared by the same method as ours has a magnetic moment of @xmath71 [ see r. h. kodama , s. a. makhlouf , and a. e. berkowitz , phys .
lett . * 79 * , 1393 ( 1997 ) ] . in the case of antiferromagnetic particles
the moment @xmath72 where n should be a number between 1 and 2 ( see ref .
@xcite and @xcite ) .
taking a value of @xmath73 , for the 5.1 nm sized particles , we estimate that , the moment should be @xmath74 . | we present a detailed study of the magnetic properties of sol - gel prepared nickel oxide nanoparticles of different sizes .
we report various measurements such as frequency , field and temperature dependence of _ ac _ susceptibility , temperature and field dependence of _ dc _ magnetization and time decay of thermoremanent magnetization .
our results and analysis show that the system behaves as a spin glass . |
the equilibrium spherical configuration of the non rotating self gravitating system is found from the set of tolman
oppenheimer volkoff ( tov ) equations @xcite , well known to be : @xmath5 here , all quantities are in usual units , @xmath6 is the pressure at radius @xmath7 , @xmath8 is the mass density of the matter which includes all forms of energy together with the rest mass , @xmath9 is the newtonian gravitational constant , @xmath10 is the light speed .
quantity @xmath11 in eq.([sdco01a ] ) is the `` mass inside radius @xmath7 '' : @xmath12 the total mass of the body , @xmath13 , is the integral of eq.([sdco02 ] ) from @xmath14 to @xmath15 .
but @xmath16 , i.e. @xmath17 .
when @xmath18 , @xmath11 must become equal to @xmath13 , so that the interior metric matched smoothly the exterior schwarzschild metric . ]
terms @xmath19 and @xmath20 after decomposition in series with respect to @xmath21 and @xmath22 of right part of eq .
( [ sdco01b ] ) , give contributions produced by the effects of the special and general theories of relativity . ] to obtain the density distribution and the mass radius relationship @xmath1 for both ordinary stars and dense compact objects , for a chosen equation of state ( eos ) @xmath23 , eqs.([sdco01b ] ) ( [ sdco02 ] ) must be integrated subject to boundary conditions @xmath24 and @xmath25 . obviously , the dependence @xmath1 is strongly model
dependent on the form of the eos a crucial point in this problem . in this paper , for the eos in our consideration
, we additionally allow for two possibilities : the possibility that @xmath26 for some @xmath27 , and the possibility of the existence of the critical state corresponding to some density @xmath2 and some temperature @xmath3 .
the realization of the latter has been supported by nuclear experiments on collisions of heavy nuclei ( see figs .
[ phasediagram ] and [ tckarn11 ] ) . in this paper
, we establish conditions in which a super dense matter ( governed by such eos ) can exist in its `` liquid '' phase in a form of stable radial distribution ( like a droplet ) .
we show also that such stable configuration is possible only for temperatures lower than the critical one .
( neutron drip ) . for densities
higher than @xmath28 , the nuclei begin to dissolved and merge together by forming the neutron liquid . in the range of higher densities ,
physical properties of the matter are uncertain . at present
, it is believed that hadronic matter , at high enough densities , undergoes a transition to a deconfined state of quarks and gluons @xcite , @xcite . in a general context
, it is assumed that as the density temperature increase above the `` normal '' nuclear characteristics , the matter may undergo phase transitions to qualitatively new states ( meson condensation , crystallization , quark deconfinement , mixed phases , etc . ) : the very existence of these states depends on the specific features of strong interactions and the quark structure of baryons .
[ phasediagram],width=264 ] by the term `` super - dense compact object '' ( sdco ) , we call a gravitationally - powerful stellar body , a remnant of a cosmic cataclysm , whose average density is of order of the nuclear density ; the density in the center of a sdco must be an order of magnitude higher than @xmath28 ] , whose mass is ( meaningfully ) smaller than the sun mass , and whose physical state in the inner region can be modeled by an eos permitting the multi phase state ( the region in fig . [ phasediagram ] where @xmath30 and @xmath31 ) .
the physics of the compact star population ( white / black dwarfs , traditional neutron stars , hyperon stars , strange stars and possible quark stars or so - called hybrid stars ) involves a complicated interplay between nuclear processes and astrophysical phenomena ( @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite ) .
the equation of state ( eos ) the dependence of pressure @xmath32 on energy ( mass ) density @xmath8 is central to the calculation of compact star s properties as it determines the mass range , the mass radius relationship , and other characteristics @xcite .
permits measuring , in principle , radius @xmath33 and allows to verify the acceptability of a particular model of eos . here
, @xmath34 is the mass of body , @xmath33 is its radius , @xmath9 is the gravitational constant , @xmath10 is the light speed . ]
the rationale for the possible form of eos has been based on data obtained from diverse sources , such as studies of high energy nuclear collisions , the monopole resonance in finite heavy nuclei , astrophysical supernovae and neutron star studies @xcite , @xcite , @xcite .
the general concept of the contemporary understanding is illustrated in fig .
[ phasediagram ] .
the critical temperature @xmath3 and the critical density of energy @xmath35 which is proportional to critical density @xmath2 , can be measured nowadays in nuclear experiments .
[ tckarn11 ] presents the various experimental data . measured by different techniques . from @xcite .
, width=264 ] obviously , of importance for sdco are the effects of the general relativity theory ( gr ) .
the significance of the effects of gravity for a body of radius @xmath33 with the inner parameter @xmath35 ( critical energy density ) , is determined by the dimensionless parameter @xmath36 ( see below ) which is small for physically interesting situations .
we take into account that the high density matter of any stable object must conform with ( i ) causality ( the speed of propagation of small matter perturbations must not exceed the speed of light ) , ( ii ) hydrodynamical stability . in sec . 2
, we describe the model which permits a radially symmetrical distribution of mass ( the choice of an equation of state , the dimensionless formulation of the set of equations , the necessary thermodynamical quantities ) .
the configuration of non rotating stars is described by the relativistic equation of hydrostatic equilibrium for a spherically symmetric body - the tov
equations @xcite . in sec . 3
, we find the matter density distribution .
the causality condition ( the speed of sound must not exceed the speed of light ) is discussed in sec .
radius dependence for the sdco is obtained in sec .
a special case is briefly considered in sec .
6 . in the concluding sec
. 7 , we discuss the obtained results .
the basic equations are written above , eqs .
( [ sdco01b])([sdco02 ] ) . the principal point for the following analysis is to propose a physically reasonable explicit form of the equation of state ( eos ) .
in fact , the exact eos of dense matter remains a well kept secret of nature in spite of decades of very intense theoretical and experimental studies ( see more detailed discussion for example the works by @xcite
. a reasonable model for the nuclear matter eos must be thermodynamically self consistent and reproduce such quantities as normal concentration @xmath37 and incompressibility factor @xmath38 in the vicinity of normal nuclear matter . at normal nuclear density
the free energy must be at minimum ( then the system is mechanically stable because of vanishing pressure @xmath39 ) and the width of this minimum is defined by the incompressibility factor @xmath38 .
the behavior of the model in other regions of the @xmath40 plane can be then probed via heavy ion collisions .
an important point is the so called ground state of nuclear matter , i.e. , at @xmath41 nuclear matter saturates ( pressure @xmath42 ) at a concentration of about @xmath43 .
the nucleon - nucleon interaction is generally attractive at nucleon nucleon separations of @xmath44 but becomes repulsive at small separations ( @xmath45 ) making nuclear matter difficult to compress . as a consequence ,
most stable nuclei are at approximately the same saturation density , @xmath46 , in their interiors .
matter at densities of up to @xmath47 may be present in the interiors of neutron stars , and matter at densities up to about @xmath48 may be present in the core collapse of type ii supernovae . in @xcite ( see also @xcite and @xcite ) , the empirical nuclear equation of state ( eos ) , polynomial approximation , @xmath49 has been proposed to explain observed experimental data . here , @xmath50 is temperature , @xmath51 , @xmath52 is energy density of the matter , @xmath53 and @xmath54 .
such eos permits the existence of the critical state when the first and second derivatives of pressure with respect to density turn zero ( surface tension vanishes ) .
coefficients @xmath55 and @xmath10 depend directly on the value of the critical temperature @xmath3 , critical density @xmath2 . despite the fact that a heavy nucleus is a `` repository '' of strongly interacting fermions , and the first term in this formula refers to the classical system , eq .
( [ n01 ] ) works well and permits to satisfactorily describe results of experiments .
this is explained by the fact that the finite system of strongly interacting fermions is satisfactorily described in terms of green s function as an ensemble of ( localized in finite volume ) collective perturbations ( of the spin zero ) which have the non zero effective `` mass '' and the degree of excitation of which can be characterized by some effective `` temperature '' .
the meaning of eq .
( [ n01 ] ) , or other interpolating expressions below , is as follows : in the gas state under normal conditions when @xmath56 , interaction between particles is very weak .
as the interaction ( pressure ) increases , the properties of the system differ more and more from the properties of the ideal gas , and finally the gas enters its condensed state liquid . in the liquid state ,
interaction between particles is great , and properties of this interaction strongly depend on the specific type of the liquid .
this is the reason why general formulae , describing quantitatively properties of liquids , do not exist ( see @xcite@xcite ) .
however , it is possible to propose some _ interpolation _ formula which can qualitatively describe the transition between the gas and liquid ( as done in the van der waals classical model ) .
such formula must produce qualitatively correct results in two limit cases . for rarified gases it should converge into formulae correct for ideal gases . but as the density increases
, it should incorporate the fact that the compressibility of the matter is limited .
such formula then would qualitatively describe the gas behavior in the transition state .
( [ n01 ] ) represents only one of the numerous possible interpolation formulae satisfying the posed requirements .
there are no physical reasons to prefer one such interpolation over the others .
but the form ( eq . [ n01 ] )
is one of the simplest and easiest to work with .
the equations of state of a multi body system of nucleons interacting via skyrme potential is presented in fig .
[ nucl - phase2 ] .
the very steep part of the isotherms ( on the left side ) corresponds to the liquid phase .
the gas phase is presented by the right parts of the isotherms where pressure is changing smoothly with increasing volume .
of special interest is the part of the diagram where the isotherms correspond to the negative compressibility , i.e. @xmath57 .
this is the so - called _ spinodal _ zone where the matter phase is unstable and can exist in both liquid and/or gas states .
liquid fog phase transition '' . ] within the spinodal zone lies a particularly unstable two phased region ( marked by the hatched line in fig .
[ spinodal - reg ] ) , in which random density fluctuations lead to almost instantaneous collapse of the initially uniform system into a mixture of two phases . for nuclear matter , it is either liquid droplets surrounded by gas of neutrons , or homogeneous neutron liquid with neutron gas bubbles ( i.e. the spinodal zone where the squire of adiabatical speed is negative , is inside the coexistence zone where the squire of isothermical speed is negative ) . for a nuclear system interacting through a skyrme potential and a van der waals compressible liquid
gas system ( shown in relative units ) .
( from @xcite.),width=245 ] phase diagram for nuclear matter ( adapted from @xcite ) .
the solid line is determined by condition @xmath58 and marks the phase transition zone .
density is expressed in units of @xmath59 .
temperature is expressed in @xmath60 units ( @xmath61).,width=264 ] critical temperature @xmath3 for the liquid - gas phase transition is a crucial characteristic of the nuclear equation of state . a typical set of isotherms for an equation of state ( eos ) - pressure versus density with a constant temperature - corresponding to nuclear interaction ( skyrme effective interaction and finite temperature of hartree
fock theory , see @xcite ) is shown in fig .
[ br08 ] .
it exhibits the maximum - minimum structure typical of the vdw like eos . depending on the effective interaction chosen and on the model ( see @xcite , @xcite , @xcite , @xcite ) , the nuclear equation of state exhibits a critical point at @xmath62 and @xmath63 ( @xcite , @xcite )
. calculations of @xmath3 were performed in @xcite , @xcite , @xcite , @xcite , @xcite , @xcite .
experimental data are presented in fig .
[ tckarn11 ] . some model eos
pressure vs. internal energy density are given in appendix [ ap:1 ] .
we introduce dimensionless variables @xmath64 .
then the dimensionless quantity @xmath65 is defined as @xmath66 and @xmath34 becomes @xmath67 we suppose that the pressure is measured in units @xmath68 , i.e. @xmath69 . then eqs . ( [ sdco01b ] ) and ( [ sdco01a ] ) for dimensionless quantities @xmath11 , @xmath70 and @xmath71 become @xmath72 here , the dimensionless parameters @xmath73 when @xmath74 , @xmath75 and @xmath76 with @xmath77 .
the system of obtained equations with an eos @xmath78 contains now only one dimensionless parameter @xmath79 which is defined by the size of the sdco , @xmath80 . if @xmath81 , we write in eqs .
( [ eos:13a ] ) and ( [ eos:13b ] ) @xmath82 and neglect further small terms of order @xmath83 ( and higher ) .
the newtonian approximation corresponds to contribution of terms @xmath84 .
the effects of relativity are taken into consideration in such approximated version of eqs .
( [ eos:13a ] ) and ( [ eos:13b ] ) terms with @xmath71 in right part of eq .
( [ eos:13b ] ) following from special relativity , and terms @xmath85 following from the gr . in the definition of the small parameter @xmath86
, @xmath9 is the gravity constant , @xmath10 is the light speed , dimensionless @xmath87 , @xmath88 .
the set of equations ( [ eos:13a ] ) ( [ eos:13b ] ) is subject to boundary conditions ( see below ) .
all principal thermodynamical quantities such as pressure , internal energy , sound speed and so on , can be calculated when a thermodynamical potential is given ( see appendix , eqs.([mod04 ] ) .
sometime , this is the free energy of the system which is preferable for description of a specific system .
to illustrate , we consider the simplest mono component system composed of @xmath89 particles occupying volume @xmath90 . following the definition , the full free energy of a system is a function of temperature @xmath50 and of volume , @xmath91 . with respect to one particle , with finite @xmath92 even when @xmath93 ( thermodynamical system ) , the fundamental thermodynamical relationship is written as @xmath94 with the chemical potential @xmath95 per one particle defined from @xmath96 .
introducing the dimensionless thermodynamical arguments @xmath97 , @xmath98 , @xmath99 , @xmath100 , we have the expressions @xmath101 the dimensionless expression for internal energy per volume unit is calculated as @xmath102 we propose the following simple expression for the dimensionless free energy @xmath103 per one particle @xmath104 here , @xmath105 are dimensionless constants , the function @xmath106 is the polylogarithm function of argument @xmath107 . in limit cases , @xmath108 for @xmath109 , and @xmath110 for @xmath111 .
the physical meaning of every term in @xmath112 and numerical values of constants @xmath113 will be discussed later .
the last term is added to set the value of entropy equal to zero at the critical point .
this additional term changes nothing in measurable quantities ( pressure , internal energy , etc ) .
the model eq.([eos : fe ] ) must produce the following results : ( a ) the eos following from eq.([eos : fe ] ) has to have a form permitting the existence of the critical point where @xmath114 ; ( b ) pressure @xmath115 when @xmath116 ; ( c ) the critical density @xmath2 is of order of @xmath117 , i.e. @xmath118 ; ( d ) compressibility factor @xmath119 ; ( e ) the principle of causality is respected - the adiabatical sound speed is always smaller than the light speed - @xmath120 .
the dimensionless pressure and volume density of internal energy are calculated now as ( see appendix [ ap:3 ] ) @xmath121 here , pressure and internal energy per volume unit contain terms which are proportional to effective temperature as for the classical ideal gas .
simple consideration shows that the free energy per one particle @xmath112 can not be a polynome of power higher then @xmath122 .
in fact , let , for large @xmath70 , the leading term in @xmath112 be proportional to some power of @xmath70 , i.e. @xmath123 with @xmath124 . since ( see appendix [ ap:3 ] and eqs . ( [ eos:06 ] ) and ( [ eos:07 ] ) ) , @xmath125 and @xmath126 with the same coefficients of proportionality , the adiabatical sound speed ( normalized by the light speed square ) @xmath127 for @xmath128 .
this gives the one possibility : @xmath129 .
obviously , it follows from here that the interpolating function in eq .
( [ eos : fe ] ) must tend to zero , when @xmath130 , not so fast as a linear function of @xmath70 , and tend to infinity not faster then @xmath131 when @xmath128 .
expression ( [ eos:07 ] ) , as well as eq .
( [ eos:06 ] ) , has a simple physical meaning : the first term @xmath132 is determined by the rest mass of particles , the second is connected with heat motion , the third describes interaction ( attraction ) of particles for moderate density of matter , the last term is connected with repulsion due to hard `` core '' inside particles .
if there are no `` wonders '' in the region of moderate densities , @xmath133 , the interpolating function can be taken in the simple ( quasi - logarithmic ) form .
the physical reasons of such comportment are the subject of more detailed and specific investigation which are outside the framework of this paper .
the constants @xmath113 are fixed by the following three conditions : ( a , b ) the first and second derivatives of the pressure @xmath134 are zero in the critical point @xmath135 ; ( c ) the dimensionless factor of incompressibility must give the experimentally obtained value .
the value of @xmath136 for different temperatures @xmath137 is found from the condition at the boundary @xmath138 .
the physical solution of this equation exists not for all values of parameters . the dependence @xmath139 , or @xmath140 , is shown in fig . [ temperature - size ] . at the boundary @xmath141 as a function of temperature @xmath137 . the spherical configuration for the liquid phase exists when @xmath142 and when @xmath143 ( the right part of the curve ) .
[ temperature - size ] ] the important parameter for any eos is the incompressibility factor @xmath38 .
it measures the stiffness of the eos , usually defined as a slope of the pressure at the point of fundamental state ( saturation point ) : @xmath144 ( in usual units ) .
this expression ( in its dimensionless form used in the paper ) is written as @xmath145 .
in fact , @xmath146 is equal to @xmath147 in the fundamental state ( when @xmath148 and @xmath149 where @xmath150 ) .
when @xmath151 , we can use @xmath152 .
the module of inelasticity is expressed in @xmath60 when @xmath3 is given in @xmath60 : @xmath153 .
the experimental value of the parameter is of order @xmath154 ( see for example @xcite ) with a value of @xmath155 ( with considerable error ) , or @xmath156 from @xcite , @xcite .
some experimental value of the incompressibility of symmetric nuclear matter at its saturation density @xmath37 has been determined to be @xmath157 @xcite .
so , the situation is still not very clear : the analysis of the sideward anisotropy observed in the heavy ions collisions at low and intermediate energies require @xmath158 , whereas the the elliptic flow anisotropy observed in the same experiment requires @xmath159 @xcite , @xcite . in the framework of our model when the critical temperature is taken as @xmath74
, we find for parameters @xmath160 numerical values @xmath161 with the factor of incompressibility @xmath162 ( i.e. @xmath163 ) , which is a satisfactory result given the approximations made along the way .
the behavior of pressure @xmath134 calculated from eqs.([eos : tqa ] ) and ( [ eos : fe ] ) , is shown in fig.[press ] for different temperatures . as a function of normalized density @xmath70 for several values of normalized temperature @xmath137 : @xmath164 ( lowest line ) , @xmath165 ( second line from bottom ) , @xmath166 ( second line from top ) with point where @xmath114 , and critical isotherm @xmath167 ( upper line ) .
all curves below the critical isotherm , i.e. when @xmath168 , possess two turning points ( @xmath169 ) where @xmath170 . in the domain @xmath171 ,
the matter is in its gas state . in the domain @xmath172 , the matter is in its liquid state . between @xmath173 and @xmath174 ,
lies the zone , where the gas and liquid phases co - exist . ]
the behavior of internal energy per volume unit is shown in figs.[enintfig2a ] and [ enintfig2b ] . as a function of normalized density @xmath70 for several values of normalized temperature @xmath137 : @xmath164 ( lowest line ) , @xmath165 ( second line from bottom ) , @xmath166 ( second line from top ) with point where @xmath114 , and the curve corresponding to critical temperature @xmath167 ( upper line ) . ] as a function of normalized density @xmath70 for large densities and for several values of normalized temperature @xmath137 : @xmath164 ( lowest line ) , @xmath165 ( second line from bottom ) , @xmath166 ( second line from top ) with point where @xmath114 , and the curve corresponding to critical temperature @xmath167 ( upper line ) . ]
fig.[pe - relationfig ] shows the pressure @xmath71 vs internal energy density @xmath52 relation for the matter in liquid phase with limit temperatures @xmath175 and @xmath176 for which the liquid state , @xmath177 , can exist . when @xmath178 , the limit curves tend to the universal relation @xmath179 . ]
the proposed model is not in contradiction with the existing models ( see fig.[fig - wnr07 ] ) .
for example , according to our model , when @xmath180 ( with @xmath74 and @xmath181 ) , we obtain for the pressure @xmath182 , i.e. the @xmath183 dependence shown in fig.[pe - relationgfig ] is near the rmf ( npkh ) model . .
the factor of transformation to usual units is @xmath184 .
[ pe - relationgfig ] ]
the adiabatical sound speed ( i.e. speed of propagation of small `` acoustical '' perturbations @xmath185 the speed of transport of energy @xmath185 the speed of transport of information ) for very high densities must not be greater than the light speed . the adiabatical ( dimensionless here ) sound speed for relativistic fluid is calculated using the expression @xmath186 .
this quantity is calculated in condition that the entropy per one particle @xmath187 is constant . however , the pressure and internal energy in the model are functions of density @xmath70 and temperature @xmath137 . therefore , it is more natural to calculate @xmath188 using the jacobians and their properties ( see @xcite , @xcite for more information ) . using the jacobians
we can find the expression @xmath189 eqs.([mod04 ] ) , ( [ mod08b ] ) with ( [ eos : fe ] ) permit calculating of all derivatives in eq .
( [ mod08b ] ) and finding @xmath188 .
dimensionless speed of sound in our model always satisfies the condition @xmath190 ( fig .
[ v2c ] ) .
the system becomes unstable with respect to small spontaneous perturbations ( fluctuations ) when @xmath191 ( fig .
[ v2b ] ) . , normalized by the speed of light , as a function of normalized density @xmath70 for several values of normalized temperature @xmath137 : @xmath192 ( upper line ) , @xmath193 ( the curve @xmath194 touches the horizontal axis in plane ( @xmath195 ) ) and @xmath164 ( lower line ) .
the domain with @xmath196 , where the sound speed @xmath197 is imaginary , is the so - called `` spinodal '' zone .
the condition @xmath196 indicates that small spontaneous initial perturbations of density will grow exponentially fast .
the development of the instability in a homogeneous medium leads to formation of a two phase configuration when liquid ( drops ) and gas ( vapor ) states co exist .
only the states corresponding to temperatures below some temperature @xmath198 ( unique for the medium ) , for which the curve @xmath199 touches the horizontal axis in plane ( @xmath200 ) , possess such domain of instability .
for the states with @xmath201 , the speed of sound is always real ( @xmath202 ) and the matter exists in a mono phase state . ] .
square of adiabatical sound speed @xmath203 as function of ( large ) @xmath70 for @xmath192 ( upper line ) , @xmath193 and @xmath164 ( lower line ) . in the proper framework
, the speed of sound can not exceed the speed of light : @xmath204 .
bethe - johnson or van der waals equations of state violate this requirement . in our model ,
the sound speed correctly tends ( at large z ) to the speed of light ( unit of one ) . ] .
the quantity @xmath205 as function of ( large ) @xmath70 in logarithmic scale . as pictured , in our model , the sound speed correctly tends ( at large z ) to the speed of light ( unit of one ) and does not exceed it . ] ) ( inside the domain , @xmath206 ; outside the domain , @xmath207 ) . ] ) ( inside the domain , @xmath206 ; outside the domain , @xmath207 ) .
the coordinates of some points in which the pressure is zero are shown : @xmath208 .
any process pushing the system from initial state ( @xmath209 ) into the spinodal region adiabatically ( line @xmath210 ) , leads to the development of collective instability and to the fragmentation of matter . ]
eqs . ( [ eos:13a ] ) , ( [ eos:13b ] ) , ( [ eos:06 ] ) and ( [ eos:07 ] )
complete the system of equations from which a radially symmetrical distribution of mass within the sdco can be found .
the set of the equations is subject to the boundary conditions @xmath211 . at this point
, there is a difference with traditional approaches when eqs .
( [ eos:13a ] ) ( [ eos:13b ] ) are numerically integrated for a given central density .
we write in eqs .
( [ eos:13a]),([eos:13b ] ) , ( [ eos:06 ] ) and ( [ eos:07 ] ) @xmath212 and @xmath213 . here , @xmath136 is the normalized on @xmath2 density at @xmath214 ( i.e. on the boundary where pressure is assumed zero ; @xmath215 ) .
quantities @xmath216 and @xmath217 may be considered as `` add - ons '' , small perturbations of the basic state .
the rough estimation of the validity of such consideration can be written as @xmath218 for the hypothetically taken @xmath219 , i.e. for @xmath220 .
the first boundary condition is obviously @xmath221 .
the second is not @xmath222 because @xmath223 : when @xmath224 , @xmath65 has to tend to @xmath225 in leading approximation .
the boundary condition for @xmath226 when @xmath227 , is chosen after solving the set of equations from condition that the quantity @xmath228 has no singularities at any point in the region @xmath229 .
thus , @xmath230 are the proper values of the set equations .
after substitution of these expressions into eq .
( [ eos:13b ] ) , and equating coefficients with the same power of @xmath231 , we obtain the set of equations for the normalized density add - on @xmath232 and mass add - on @xmath233 which is resolved by the standard procedure .
the density distribution for the hot sdco with @xmath234 , is shown in figs.([zs - config ] ) and ( [ ldh ] ) . ) as a function of distance from center @xmath187 for different values of the sdco radii @xmath80 : @xmath235 ( lower line ) , @xmath236 and @xmath237 ( upper line ) . ] ) as a function of distance from center @xmath187 for large values of @xmath80 : @xmath238 ( lower line ) , @xmath239 and @xmath240 ( upper line ) .
for such great density , the density distribution is quasi - homogeneous and is essentially defined by the gr effect . ]
the mass radius relationship follows from eq .
( [ eos:10a ] ) . to simplify the final expressions
, we introduce special units : the mass @xmath34 of a sdco will be measured in @xmath241 units and the radius , @xmath242 , is @xmath80 units . here , the sun s mass and radii are denoted by @xmath243 .
numerically , @xmath244 the mass of the sun @xmath245 , @xmath246 , @xmath247 , @xmath248 .
the mass radii relationship becomes @xmath249 it can be also presented in the form @xmath250
a^3 f(a , \theta , z[1]).\end{aligned}\ ] ] ) $ ] defining the mass radius relationship ( in @xmath251 units for mass and @xmath252 for size ) for cold ( @xmath148 ) and hot ( @xmath253 ) sdco .
the domain where @xmath254 corresponds to the case when one can neglect the gr gravity effect .
the thin dotted lines correspond to the newtonian approximation . ]
mass radius relationship @xmath255 ( mass in @xmath251 units ) for the considered special case when the temperature @xmath256 , i.e. near the critical temperature @xmath257 , is shown in fig .
[ mass - rad ] .
units for mass and @xmath252 for size ) for a hot sdco ( with temperature @xmath234 ) . dashed line plots the relationship in the newtonian approximation . ]
we considered conditions for which sdcos small stellar bodies with a specific nuclear equation of state can exist in the nature .
the sdco s mass as a function of its size , has been obtained .
the principal point of the work is the interpolating expression for the dimensionless free energy ( eq . [ eos : fe ] ) from which all thermodynamical quantities can be found .
the expression for the free energy reflects the following limit conditions : for small densities , @xmath130 , the interaction between particles is weak , and the dominant term is the first term which describes a gas of non - interacting particles . as the density increases , the properties of the system differ more and more from the properties of the ideal gas , the interaction ( logarithmic term in expression for pressure ) becomes more and more significant . with more increasing of density , @xmath178 , the gas enters its condensed state ( liquid ) when the term @xmath258 in expression for @xmath112 is the most important . for high densities
@xmath70 , the equation of state has to be `` hardened '' to account for the dominance of the `` repulsive core '' in the potential of particle interaction . in such `` hardened '' state , repulsion between particles
is very strong , and the properties of this interaction no longer depend on the specific type of the liquid , thus the corresponding term in the free energy has to have a universal form for the pressure @xmath259 corresponding to arguments of @xcite .
the proposed model eq.([eos : fe ] ) responded to the following requirements : ( a ) the eos following from eq.([eos : fe ] ) has to have a form admitting the existence of the critical point where @xmath114 ; ( b ) the pressure @xmath115 for some value @xmath116 ; ( c ) the critical density @xmath2 is of order of @xmath117 , i.e. @xmath118 ; ( d ) compressibility factor @xmath119 ; ( e ) the principle of causality is respected - the adiabatical sound speed is always smaller than the light speed . the proposed model of the eos permitted to construct a spherical self gravitating configuration : the sdco . in figs . [ press ] - [ mass - rad ]
we graphically illustrate some of the obtained results : the equations of state @xmath260 and the square of sound speed @xmath261 for various values of temperature @xmath137 ( and scale zooms for @xmath70 ) , resulting from the model ( eq . [ eos : fe ] ) .
the calculations have been made with parameters @xmath262 .
the figures demonstrate the existence of the spinodal zone for temperatures below critical , where the square of the sound speed is negative .
this signifies that the speed of sound is imaginary in the domain , indicating that small spontaneous initial perturbations of matter density grow exponentially fast in beginning of process .
the instability process leads to formation of the liquid gas phase state .
the model correctly captures the principle of causality when the speed of propagation of small perturbation of matter density is smaller then the light speed .
any process pushing the system from initial `` liquid '' state ( @xmath209 ) into the spinodal region for example adiabatically ( following to lines @xmath210 ) , leads to instability development and fragmentation . to obtain such a situation ,
it is sufficient to `` rarify '' some domain of the sdco ( for example @xmath263 ) .
it can be accomplished by a sharp deceleration of the sdco .
a stationary spherical configuration exists only if the boundary condition for pressure @xmath264 is respected for some @xmath116 . with respect to fig . [ press ] , this signifies that there is intersection of curves @xmath265 for given @xmath266 with horizontal axis @xmath264 .
the @xmath116 is the boundary value of density which corresponds to @xmath267 . if some mechanism ( for example due to a simple deceleration of the object colliding with another massive object ) introduces some quantity of heat into the sdco , the system passes into a new state characterized by a new value of temperature , @xmath268 . in this case
, the curve of @xmath269 may not intersect the horizontal axis @xmath264 : instead , for example , of the middle or lowest lines in fig . [ press ]
, the configuration will be characterized by the upper line .
this signifies that an equilibrium spherical configuration for the sdco does not exist more : all is manifested as an explosion of the system and its destruction in multitude of fragments which can be unstable too with respect to specific nuclear reactions .
the model admits a generalization to a multi component system , including thermic radiation . for this , the expression for the free energy must be replaced by the sum of the free energy expression for every component .
a number of models for the eos of neutron matter have been presented in literature over the years . according to @xcite where the references are presented
, these models can roughly be classified as follows : thomas - fermi based models ; schroedinger - based models ( e.g. variational approach , monte carlo techniques , hole line expansion ( brueckner theory ) , coupled cluster method , green function method ) ; relativistic field - theoretical treatments ( relativistic mean field ( rmf ) , hartree - fock ( rhf ) , standard brueckner - hartree - fock ( rbhf ) , density dependent rbhf ( dd - rbhf ) ; nambu - jona - lasinio ( njl ) models ; chiral su(3 ) quark mean field model . a collection of equations of state computed for several of these models is shown in fig .
[ fig - wnr07 ] .
$ ] meson condensate ; `` q '' signifies @xmath270 quarks ; `` h - matter '' is the h - dibaryon condensate .
[ fig - wnr07 ] ] all presented here models show a monotone dependence of pressure as function of energy density in the region of moderate values , and the two phase state of matter is not realized in framework of the models .
the fundamental thermodynamic relationship ( ftr ) for the change of the internal energy @xmath271 of a system is @xmath272 . here , all quantities have the standard meaning and dimension , @xmath50 is the temperature of the system , @xmath273 is the entropy , @xmath32 is pressure , @xmath90 is the full volume , @xmath274 is the chemical potential , and @xmath89 is full number of particles .
it follows from here that @xmath275 . for the free energy @xmath276 , we obtain from the first law of thermodynamics that @xmath277 . from here
it follows that the free energy is a function of temperature , volume , and number of particles : @xmath278 .
if we introduce quantities reported to one particle , @xmath279 , @xmath280 , @xmath281 , @xmath282 where @xmath13 is the full mass of the system , @xmath90 is its volume , @xmath283 is the nucleon mass , we can write that the free energy reported to one particle is @xmath284 and its differential is @xmath285 .
the pressure is thus a function of temperature and particle concentration : @xmath286 .
the chemical potential and the internal energy also per one particle are @xmath287 and @xmath288 .
the internal energy reported to volume unit ( which has the same dimension as pressure ) is given by @xmath289 .
for the _ dimensionless _ quantities defined by expressions @xmath290 , we can write @xmath291 pressure @xmath71 , entropy @xmath187 and the volume density of internal energy @xmath52 are calculated as @xmath292 note , if the free energy @xmath112 is given as a function of density @xmath70 and temperature @xmath137 , the other thermodynamical quantities depend on @xmath70 and @xmath137 too . in some situations ,
the so - called grand canonical potential , @xmath293 , also called the landau free energy , or landau potential , is more useful .
the quantity @xmath293 is defined via the full free energy @xmath294 of the system as @xmath295 .
the change of the grand potential is found from definition of @xmath293 and the ftr and is given by @xmath296 .
this expression shows that the grand canonical potential is a function of parameters @xmath297 : @xmath298 . letting @xmath299 , @xmath300 and @xmath301
, we obtain that @xmath302 .
it follows from here that , for homogeneous thermodynamical systems , @xmath303 , or @xmath304 and @xmath305 @xcite .
the density of particles is given by derivative @xmath306 , the entropy reported to the unit of volume is @xmath307 and the internal energy ( in terms @xmath308 ) reported to volume unit is @xmath309 .
the combination @xmath310 gives a volume density of generalized free energy @xmath311 with @xmath312 which is equally expressed in terms @xmath308 : the function @xmath313 has the same dimension as pressure @xmath32 .
resolving @xmath314 , one finds @xmath313 in traditional form @xmath315 akmal , a. , pandharipande , v.r .
& ravenhall , d.g .
c * 58 * , 1804 - 1828 ( 1998 ) shen , h. , toki , h. , oyamatsu , k. & sumiyoshi , k. : nucl . phys .
a * 637 * , 435 - 450 ( 1998 ) douchin , f. & haensel , p. : astron . astrophys . * 380 * , 151 - 167 ( 2001 ) karnaukhov , v.a .
particles * 37 * , 2 , 313 ( 2006 ) | the equation of state ( eos ) @xmath0 pressure as a function of density and other thermodynamical quantities is what generates particularities of mass radius distribution @xmath1 for super dense compact stellar bodies , the remnants of cosmic cataclysms . in view of recent nuclear experiments ,
we propose one particular eos , which admits the critical state characterized by density @xmath2 and temperature @xmath3 , and which under certain conditions permits a radial distribution of the super dense matter in `` liquid '' phase .
we establish such conditions and demonstrate that a stable configuration is indeed possible ( only ) for temperatures smaller than the critical one . using tolman oppenheimer
volkoff equations for hydrostatic equilibrium , we derive the mass radius relation for the super dense compact objects with masses smaller than the sun , @xmath4 .
the obtained results are within the constraints established by both heavy
ion collision experiments and theoretical studies of neutron rich matter . |
energy harvesting is the process by which energy is taken from the environment and transformed to provide power for electronics.@xcite specifically , thermoelectrics can play a crucial role in future developments of alternative sources of energy .
unfortunately , present thermoelectric engines have low efficiency.@xcite therefore , an important task in condensed matter physics is to find new ways to harvest ambient thermal energy , particularly at the smallest length scales where electronics operate .
utilizing the physics of mesosopic electron transport for converting heat to electrical power is surprisingly a relatively recent endeavor .
while the general relationships between electrical and heat currents and their responses to applied voltages and temperature differences have been understood since the work of onsager,@xcite the investigation of thermoelectric properties , and in particular the design of nano - engines has taken the form of several fairly recent concrete proposals . in 1993 ,
hicks and dresselhaus investigated the thermoelectric properties of a mesoscopic one - dimensional wire.@xcite mahan and sofo subsequently showed that the best energy filters are also the best thermoelectrics.@xcite this suggests the use of quantum dots with discrete energy levels to investigate thermodynamic questions.@xcite the seebeck effect - the appearance of a voltage when there is a temperature difference across a sample - was investigated for a single quantum dot with a resonant level by nakpathomkun _
et al._@xcite resonant levels were also used as energy filters to make a related heat engine from an adiabatically rocked ratchet.@xcite humphrey _ et al .
_ note their model can be generalized to a static , periodic ratchet , which is a quantum version of the model with state - dependent diffusion.@xcite references show that a single resonant level is an ideal heat engine , and investigate power and efficiency in that system which has similarities to the present work . .
the cavity is kept hot through coupling to an energy reservoir at temperature @xmath0 ( not shown ) which is considered larger than the reservoir temperature @xmath1 .
thermal broadening of the fermi functions in the three regions ( source , cavity and drain ) is shown by the light shading .
( a ) rectification configuration . in the absence of bias ( short circuit ) , electrons enter the cavity via the left lead , gain energy @xmath2 from the cavity , and exit through the right lead , transferring an electrical charge @xmath3 through the system . ( b ) carnot efficient stopping ( open circuit ) configuration . ]
coulomb - blockaded dots can be ideally efficient converters of heat to work , both in the two - terminal @xcite and three - terminal@xcite cases ; however , since transport occurs through multiple tunneling processes , the net current and power are very small . in light of the small currents and power produced by coulomb - blockaded quantum dots , open cavities with large transmission that weakly changes with incident electron energy
have been considered.@xcite while this system produces more rectified current than coulomb - blockaded quantum dots , simply increasing the number of quantum channels does not help because the energy dependence of transmissions in typical mesoscopic conductors is a single - channel effect even for a many - channel conductor .
hence , the rectified current and power drop as channel number is increased unless special engineering of the contacts is made .
consequently , we should optimize the rectified current and electrical power for a strongly non - linear system , operating in the single channel limit .
we do this here with the physics of resonant tunneling through barriers connected to the hot energy source .
we show this achieves the maximum current a single channel can give , which we conjecture is the optimal configuration .
importantly , in our model the energy source is separate from the electrical circuit , as is required for energy harvesting .
therefore no charge is extracted from it and the optimal configuration is independent of the parameters of the energy source .
this is different from single quantum dot proposals @xcite where the leads are held at different temperatures , and the amount of transferred heat depends on the lead chemical potentials .
resonant tunneling is a quantum mechanical effect , where constructive interference permits an electron tunneling through two barriers to have unit transmission .
this is only true if the electron has a particular energy equal to the bound state in the quantum dot , or within a range of surrounding energies , whose width is the inverse lifetime of the resonant state.@xcite electrons with any other energy are effectively forbidden from transmitting across the quantum dot . in this way ,
a resonant tunneling barrier acts like an energy filter : only those electrons that match the resonant condition are permitted to pass .
we assume for simplicity the resonant tunnel barrier ( or the dot ) is symmetrically coupled .
the geometry we consider is related to one that has already been fabricated experimentally , but considered for a different purpose , based on the theoretical proposal of edwards _
et al._.@xcite prance _
et al . _ performed experiments on this system in its dual role as an electronic refrigerator.@xcite they demonstrated that applying bias to the system results in cooling a large cavity from to below .
we are primarily interested in cavities made as small as possible while still having good thermal contact with the heat source .
this further miniaturization permits many engines to be put in parallel and give a large output power .
the paper is organized as follows . in sec .
[ model ] we describe the transport through our device , the results of analytical and numerical calculations are presented in sec .
[ results ] .
section [ layered ] investigates the scaling of the simple system consisting of two quantum dots up to a layered structure with many channels contributing at the same energies and discusses the robustness of the proposed structure to fluctuations in the fabrication process .
the conclusions are presented in sec .
[ conclusions ] .
the model we consider ( shown in fig .
1 ) consists of a cavity connected to two quantum dots , each with a resonant level of width @xmath4 and energy @xmath5 .
we consider the situation where the widths are equal , while the energy levels are different ( these are controlled by gate voltages ) .
the energy difference @xmath6 is an important energy scale of our composite system , which we refer to as the _ energy gain_. it is distinct from the level spacing @xmath7 in the individual dots as well as from the level width .
the nano - cavity the dots are connected to is considered to be in equilibrium with a heat reservoir of temperature @xmath0 that is hotter than the left and right electron reservoirs , having chemical potentials @xmath8 and equal temperatures , @xmath1 .
we assume strong electron - electron and electron - phonon interactions relax the electron energies as they enter and leave the cavity , so the cavity s occupation function may be described with a fermi function , @xmath9)$ ] completely characterized by a cavity chemical potential @xmath10 and temperature @xmath0 , with @xmath11 being the boltzmann constant . this process of inelastic energy mixing is assumed to occur on a faster time scale than the dwell time of an electron in the cavity .
thermal energy flows from the coupled hot bath into the cavity as a heat current , and keeps the temperature different from that of the electron reservoirs .
the nature of the heat reservoir is not specified in this model , but refers quite generically to any heat source we wish to harvest energy from . our setup should be contrasted with the more widely studied two - terminal configurations in which electric and heat currents flow in parallel .
our model permits a separate heat circuit to the central reservoir and a separate ( transverse ) electrical circuit .
the chemical potential of the cavity and its temperature ( or equivalently , the incoming heat current ) are constrained by conservation of global charge and energy .
these constraints are given by the simple equations , @xmath12 , and @xmath13 in the steady state , where @xmath14 is electrical current in the left or right contact , and @xmath15 the energy current .
energy current is seemingly not conserved because of the heat current @xmath16 flowing from the hot reservoir .
the currents @xmath17 , @xmath18 , are given by the well known formulas @xmath19 $ ] and @xmath20 $ ] , where @xmath21 is the transmission function of each contact for each incident electron energy @xmath22 . in our quantum dot geometry , the resonant levels give rise to a transmission function of lorentzian shape,@xcite t_j(e)= , where @xmath23 are the attempt frequencies of the two barriers of the resonant quantum dot(s ) . here
we assume symmetric coupling for simplicity , @xmath24 , so @xmath4 is the width of the level , or inverse lifetime of an electron in the dot .
note that the lorentzian energy dependence applies if the level width @xmath4 is small compared to the level spacing @xmath7 in the individual quantum dots .
a crucial advantage of our setup which we will exploit later on is that it permits the use of small dots with a level spacing that is large compared to temperature but with the energy gain and level width of each dot on the order of the temperature . in the limit where the width of the level is smaller than the thermal energy in the cavity / dot system , @xmath25
, the transmission will pick out only the energies @xmath26 or @xmath27 in the above energy integral expressions for the currents giving simple equations . consequently
, we have these equations for the conservation laws for charge and energy : @xmath28 + e_r [ f_r - f_{\text{c}r } ] , \label{conservation2}\end{aligned}\ ] ] where @xmath29 , @xmath30 , @xmath31 , and @xmath32 . from these two equations
, we can solve for ( say ) the quantity @xmath33 .
this quantity is proportional to the electrical current through the left lead @xmath34 , the net current flowing through the system . a solution of eqs . and to linear order in the deviation of the cavity s temperature and chemical potential from the electronic reservoirs indicates that the maximal power of the heat engine will be produced when the chemical potentials of the reservoirs are symmetrically placed in relation to the average of the resonant levels , @xmath35 .
for this special case , an exact solution is possible because the constant solution @xmath36 for the cavity chemical potential satisfies the charge conservation condition for all temperatures .
we now describe the physics of this nano - engine .
we first focus on the regime @xmath37 , which can be analyzed analytically and will afterwards discuss the regime @xmath38 which we numerically find to yield the largest current and power . physically ,
if an electron comes in the left lead at energy @xmath26 and exits the right lead with energy @xmath39 , it must gain precisely that energy difference @xmath40 .
thus , in the steady state , any incoming heat current @xmath16 must be associated with an electrical current @xmath41 , with a conversion factor of the energy gain , @xmath42 , to the quantum of charge , @xmath3 , i = .
this results holds regardless of what bias is applied or what the temperature is .
the efficiency of our heat engine , @xmath43 , is defined as the ratio of the harvested electrical power @xmath44 to the heat current from the hot reservoir , @xmath16 . for our system
it takes a particularly simple form , = .
[ efficiency ] in order to proceed further , we must find the chemical potential of the cavity and its temperature given in terms of the incoming heat current and chemical potentials and temperature of the electron reservoirs .
these are found by employing the principle of conservation of global charge and energy , see eqs . and . for @xmath45 and level width @xmath4 and
@xmath46 optimized to give maximum power .
( b ) scaled maximum power as a function of @xmath47 for optimized values of @xmath48 and @xmath46 .
( c ) efficiency at maximum power for the values of @xmath49 and @xmath46 chosen to maximize power .
( d ) the optimized values are plotted versus @xmath47 . here , @xmath50 , @xmath51 .
] rather than considering the cavity temperature @xmath0 as a function of the heat current @xmath16 we are harvesting , we can turn our perspective around , and keep the cavity temperature @xmath0 fixed from being in thermal equilibrium with the hot energy source . with this insight
, we can express the heat current @xmath16 in terms of the hot cavity temperature and other system parameters .
we find in the limit where @xmath52 , j = [ f(e/2 , t _ ) - f(e/2 - /2 , t _ ) ] , [ jsol ] which satisfies charge and energy current conservation . in eq .
( [ jsol ] ) , @xmath53 is planck s constant . importantly , without bias there is a rectified electrical current given by i = e j / e [ ( k_b t_)^-1 - ( k_b t_)^-1 ] , [ isol ] in the limit where @xmath54 . this current is driven solely by the fixed temperature difference between the systems .
we note both the heat and electrical current are proportional to @xmath4 , the energy width of the resonant level .
consequently , the currents and power produced in this system will tend to be small since we have assumed that @xmath4 is the smallest energy scale .
it is also clear that both are controlled by the size of @xmath42 , so increasing this energy gain will improve power until it exceeds the temperature .
later we will generalize these results by numerically optimizing the power produced in this nano - engine . in order to harvest power from this rectifier ,
a load should be placed across it .
equivalently , we could apply a bias @xmath55 to this system tending to reduce the rectified current . at a particular value , @xmath56 ,
the rectified current vanishes , giving the maximum load or voltage one could apply to extract electrical power at fixed temperatures @xmath57 .
this value is found when @xmath16 and @xmath41 vanish , given by eq . : _
stop = e ( 1 - ) .
[ stop ] consequently , the voltage applied must not be larger than @xmath58 and therefore from eq .
( [ efficiency ] ) , the efficiency is bounded by @xmath59 . at the stopping voltage , the thermodynamic efficiency attains its theoretical maximum , the carnot efficiency , @xmath60 , showing this system is an ideal nanoscale heat engine .
naturally , at this point [ see fig .
[ fig1](b ) ] the system is reversible with no entropy production . also interesting
is the efficiency at the bias point where power is maximum . for temperature
larger than @xmath42 or @xmath61 , we can approximate the fermi functions to find @xmath62 , resulting in a parabola as a function of @xmath63 , with maximum power p_max , [ pmax ] and efficiency @xmath64 which is in agreement with general thermodynamic bounds for systems with time - reversal symmetry.@xcite one can go beyond this limit for the efficiency by solving the conservation laws numerically .
we optimize the total power produced by the heat engine by varying the resonance width @xmath4 , as well as the energy gain @xmath42 and applied bias @xmath65 , given fixed temperatures @xmath57 .
these results are plotted in fig .
[ fig2 ] , where we define the average temperature @xmath66 , and its difference , @xmath67 . in fig .
2(a ) , we see that for @xmath68 the power increases as @xmath69 , as indicated in eq .
( [ pmax ] ) , but then levels off and decays exponentially , attaining its maximum around @xmath70 .
similarly , the choice @xmath71 gives optimal power .
we emphasize that @xmath42 and the level width are two essentially independent energy scales in our system . as a matter of fact the energy gain
@xmath72 is almost an order of magnitude larger than the level width @xmath73 .
these considerations suggest an experimental strategy for maximizing the power of such a device : measure what the resonant level widths are , and tune the reservoir temperatures and energy gain ( with the help of gate voltages ) , to it . from fig .
[ fig2](b ) we also see that the efficiency at maximum power drops from half the carnot efficiency to about 0.2@xmath60 when the parameters are optimized .
however , we note that when @xmath4 is kept small , in the nonlinear regime the efficiency can exceed the bound @xmath74 found in the linear regime .
this small drop in efficiency is more than compensated by the extra power we obtain .
importantly , according to fig .
[ fig2 ] , the power reaches a maximum of @xmath75 , or about at @xmath76= , a two order of magnitude increase from a weakly nonlinear cavity .
@xcite this jump in power can be attributed to the highly efficient conversion of thermal energy into electrical energy by optimizing both the level width and energy level difference .
compared to a heat engine based on resonant tunneling through a single quantum dot in a strictly two - terminal geometry,@xcite our three - terminal energy harvester gains a factor of 2 in power while achieving the same efficiency . to give some perspective on this output ,
if a square array of these nanoengines were fabricated , each occupying an area of , they would produce a power of , operating at @xmath76 = .
$ ] leaving ( blue ) or entering ( red ) the central cavity vs temperature difference ( x axis ) and applied bias ( y axis ) divided by average temperature .
the plot is for system parameters optimized for maximal power output , an energy gain of @xmath77 and level width @xmath78 .
the green line is the @xmath79 curve for @xmath80 while the black curve is the @xmath79 line for the optimized @xmath4 .
the system can work as a heat engine ( he ) in the blue region for @xmath81 ( configuration shown in the inset labeled he ) , or as a refrigerator ( r ) in the blue region for @xmath82 ( configuration shown in the inset labeled r ) .
[ heatcurr ] ] in fig .
[ heatcurr ] , the heat current @xmath16 is plotted versus temperature difference and applied bias . there
we find that when system parameters are optimized to give maximum power , the system can be operated in the mode of a heat engine ( he ) or a refrigerator ( r ) .
however , in contrast to the case where the levels are narrow compared to the other energy scales [ and consequently the cavity can cool to arbitrarily low temperatures in principle ; see green solid line , eq .
] , for this choice of parameters the cavity will only cool to the temperature where the @xmath79 curve ( solid black line ) bends back . ] in reality , there will be other quantum dot resonant levels the electron can occupy that are higher up in energy .
we have assumed that the cavity temperature and applied bias are sufficiently small so transport through these levels can be neglected .
our model is quite general , so it may be applied to both semiconductor dots in two - dimensional electron gases,@xcite as well as three dimensional metallic dots.@xcite this latter case is quite interesting since one can fabricate an entire plane of repeated nanoengines in parallel in order to scale the power.@xcite furthering this idea , for such a repeated array of cavities and quantum dots , one can connect all the cavities to make a single engine , see fig .
[ swiss ] .
the two boundary layers consist of planes of quantum dots , so electrons can only penetrate through them .
these layers sandwich a hot interior region and separate it from the left and right cold exterior contacts .
this is equivalent to taking a large cavity with two leads and scaling the power by adding more quantum dots ( rather than trying to add more channels to a single contact ) .
interestingly , the layered structure can help to reduce phononic leakage heat currents that would otherwise reduce the efficiency of the system : phonons scatter efficiently at interfaces , and the random dot arrangement will further reduce phonon related heat losses .
the sandwich engine fabricated with self - assembled quantum dots tolerates variations in width and fluctuations in energy levels , as we will see in the next section .
here , we further investigate the self assembled quantum dot layered engine , and what some of its theoretical characteristics are under realistic fabrication conditions . the basic operating configuration for the engine is shown in fig .
[ swisscheese ] .
heat flows from the hot energy source we wish to harvest energy from into the engine , and is converted into electrical power , along with residual thermal energy , dumped into the cold temperature bath .
the electrical current is carried by the cold thermal bath , where it powers a load and then completes the electrical circuit on the opposite cold terminal .
the flow of heat out of the hot energy source will consequently tend to cool the hot source .
the energy harvesting application we primarily have in mind is taking heat away from computer chips and running other devices on the chip itself .
in electrical chips , thermal energy is an abundant and free resource .
indeed , heat is not only free , it is a nuisance preventing further improvements on chip technology .
the fact that the proposed heat engine not only harvests the thermal energy , converting some of it into electrical power , but also cools the hot source is therefore a side benefit to the proposal of including nanoscale heat engines as part of an emerging chip technology . in the actual fabrication of a resonant tunneling nano - engine ,
as long as there are only a few dots , the precise placement of the resonant levels can be controlled by gate voltages in order to maximize the power generated by the engine . however , as soon as we consider self - assembled quantum dots with charging energies and single - particle level spacings of the order of , @xcite thereby allowing the nano - engine to operate at room temperature , this kind of control is out of the question . to make such an engine ,
there are several possible fabrication techniques that could be employed using layers of quantum dots and wells to have the resonant energy levels lower than the fermi energy on one side of the heat source , and higher on the other side .
however , in all of these fabrication methods , the growth of quantum dots does not occur at a perfectly regular rate , so it is natural to expect there will be variation of the resonant energy level from dot to dot .
we must then check whether this fact will degrade the performance of the engine , and if so by how much . to answer this question
, we consider the energy - resolved current arising from a number of electrons passing through @xmath83 quantum dots on the left and then the right layer .
the total current coming from the left slice is given by @xmath84 , \nonumber \\
t_{{\rm eff } , l } & = & \sum_{i=1}^n t_i(e , e_i),\end{aligned}\ ] ] with similar equations for the total right current , as well as the energy currents . here
, @xmath85 is the transmission probability of quantum dot @xmath86 , which has a resonant energy level @xmath87 and a width @xmath88 , @xmath89 $ ] for symmetrically coupled quantum dots .
since neither the left nor cavity fermi functions depend on the level placement , the sum over the quantum dots can be done to give an effective transmission function @xmath90 for the whole left slice .
we can make further progress by assuming the fabrication process can be described as a gaussian random one , where the energy level @xmath87 is a random variable with an average of @xmath26 and a standard deviation of @xmath91 . for simplicity , we only consider random variation in @xmath87 , but there will also be variation in @xmath88 we ignore for the present . with this model
, the effective transmission will have the average value , the average energy gain between left and right dots .
( a ) the power is plotted for different level widths of the dots , where @xmath92 .
( b ) the power is plotted for different applied voltages , where @xmath93 . ]
t_eff , l = n t_i(e , |e ) _ p = n d|e t(e - |e ) p_g(|e , ) , where @xmath94 is the gaussian distribution described above .
thus , we see the effective transmission function is simply a convolution of the lorentzian transmission function and the gaussian distribution , known as a voigt profile .
this leads to further broadening of the lorentzian width . with these considerations ,
the conservation laws for charge and energy retain the same basic form as described in eqs .
( [ conservation1 ] ) and ( [ conservation2 ] ) , but with @xmath83 times the voigt profile playing the role of the energy - dependent transmission for the left and right leads . we have numerically solved these equations and plotted the maximum power per nanoengine versus the width of the gaussian distribution in fig . [ ensemblepower ] .
the parameters are chosen so as to optimize the engine s performance without any randomness in the level position .
as the randomness of the level position is increased , the power begins to drop as expected .
however , even when the scatter of the energy levels is @xmath95 of @xmath42 , the power only drops to @xmath96 of its maximum , showing that this engine is robust to these kinds of fluctuations in fabrication . notice that if the level width is less than the optimal amount , some disorder in the level energy can actually _ improve _ performance in comparison to a cleanly optimized level width .
this is because of the additional broadening in energy space the level disorder provides .
another interesting effect is shown in fig .
[ contourpower ] , where the value of @xmath4 is not optimal together with a given amount of disorder in the energy level positions .
the figure demonstrates that even in this experimentally realistic case , a change of the voltage and relative level spacing can yield results which are nearly as good as the optimal case .
, in units of the optimal power , @xmath97 .
the level width is taken to be a suboptimal value , @xmath98 and the width of gaussian random distribution is fixed at @xmath99 .
the subscript opt " refers to the parameter that is optimized for maximal power in the clean case . ]
in this paper , we have opened a route to highly efficient solid state energy harvesting .
our work shows that the most efficient structure is found with energy filters that have transmission probabilities close to 1 in a configuration where the level structure of spatially separated quantum dots can be independently adjusted to give the desired energy gain .
this yields a high power heat engine with no moving parts .
we show that nano - scale quantum dots can be employed in a parallel configuration that delivers substantial power .
this work was supported by the us nsf grant no .
dmr-0844899 , the swiss nsf , the nccr manep and qsit , the european strep project nanopower , the csic and fse jae - doc program , the spanish mat2011 - 24331 and the itn grant no .
234970 ( eu ) .
a.n.j . thanks m.b . and the university of geneva for hospitality , where this work was carried out . | we propose a nanoscale heat engine that utilizes the physics of resonant tunneling in quantum dots in order to transfer electrons only at specific energies . the nanoengine converts heat into electrical current in a multiterminal geometry which permits one to separate current and heat flows . by putting two quantum dots in series with a hot cavity , electrons that enter one lead
are forced to gain a prescribed energy in order to exit the opposite lead , transporting a single electron charge .
this condition yields an ideally efficient heat engine .
the energy gain is a property of the composite system rather than of the individual dots .
it is therefore tunable to optimize the power while keeping a much larger level spacing for the individual quantum dots . despite the simplicity of the physical model ,
the optimized rectified current and power is larger than any other candidate nano - engine . the ability to scale the power by putting many such engines into a two - dimensional layered structure gives a paradigmatic system for harvesting thermal energy at the nanoscale .
we demonstrate that the high power and efficiency of the layered structure persists even if the quantum dots exhibit some randomness . |
in 1961 zaitsev @xcite supposed that all solvable lie algebras of a fixed dimension could be obtained via contractions from the semisimple algebras of the same dimension .
he called such solvable lie algebras ` limiting solvable ' and proposed the contraction procedure for the ` limiting classification ' of solvable lie algebras .
later the same conjecture was also formulated by other scientists , e.g. celeghini and tarlini .
complexity of the actual state of affairs was illustrated in @xcite , where contractions of real and complex low - dimensional lie algebras were studied .
the incorrectness of the above conjecture was illustrated by the fact that all semisimple ( and reductive ) lie algebras are unimodular and any continuous contraction of a unimodular algebra necessarily results in a unimodular algebra .
moreover , semisimple lie algebras exist not in all dimensions . the aim of this paper is to revisit zaitsev s conjecture in terms of @xmath0-expansions .
we will study @xmath0-expansions of the real three - dimensional lie algebras .
it was shown in @xcite that generalized inn
wigner contractions give a particular case of @xmath0-expansions .
this is why we consider only pairs of such algebras that are not connected by a contraction .
the paper is arranged as follows .
section 2 contains preliminary information on @xmath0-expansions and relevant objects . in section 3
we construct several key examples of three - dimensional @xmath0-expansions and discuss the possibility of application of @xmath0-expansions to the classification of solvable lie algebras .
roughly speaking , the _ @xmath0-expansion _ is the `` product '' of the semigroup and the lie algebra with a lie algebra structure defined in a special way .
the notion was introduced in @xcite .
it generalizes the notions of expansion , deformation ( under a proper choice of the corresponding semigroup ) , extension and generalized inn
wigner contraction .
= -1 let @xmath1 be an abelian semigroup of order @xmath2 and let @xmath3 be an @xmath4-dimensional lie algebra with the structure constants @xmath5 in a fixed basis @xmath6 of the underlying vector space @xmath7 .
here and in what follows the indices @xmath8 , @xmath9 and @xmath10 run from 1 to @xmath2 and the indices @xmath11 , @xmath12 and @xmath13 run from 1 to @xmath14 .
the @xmath15-dimensional @xmath0-_expanded lie algebra _ @xmath16 , the underlying vector space of which is spanned by the basis elements @xmath17 , is defined by the structure constants @xmath18 [ nesterenko : basis change ] it directly follows from ( [ nesterenko : cijk ] ) that the product of elements @xmath19 and @xmath20 in the algebra @xmath21 is zero if @xmath22=0 $ ] .
therefore , an appropriate representative from the isomorphism class of the initial lie algebra @xmath3 should be chosen if @xmath0-expansion is used for a certain purpose .
in other words , sometimes the structure constants tensor has to be transformed by an element of the group @xmath23 before the expansion procedure .
of course , commutation relations of the expanded algebra @xmath21 can also be reduced to an appropriate form using a change of basis .
let a lie algebra @xmath24 admit the decomposition @xmath25 with @xmath26\subset \hat{v}$ ] , namely @xmath27=\hat{c}_{\check{i}\hat{j}}^{\hat{k } } \hat{v}_{\hat{k } } , \quad [ \check{v}_{\check{i}},\check{v}_{\check{j}}]=\check{c}_{{\check{i}}{\check{j}}}^{\check{k } } \check{v}_{\check{k}}+\hat{c}_{{\check{i}}{\check{j}}}^{\hat{k}}\hat{v}_{\hat{k } } , \quad [ \hat{v}_{\hat{i}},\hat{v}_{\hat{j}}]=\check{c}_{\hat{i}\hat{j}}^{\check{k } } \check{v}_{\check{k}}+\hat{c}_{\hat{i}\hat{j}}^{\hat{k } } \hat{v}_{\hat{k}};\ ] ] where @xmath28 is a basis of @xmath29 and @xmath30 is a basis of @xmath31 .
then the elements @xmath28 form a well - defined _ reduced _ lie algebra @xmath32 with the lie product defined by the structure constants @xmath33 .
in general case the dimension of the @xmath0-expanded lie algebra @xmath21 is greater than @xmath4 and it has a more complicated structure than @xmath3 .
therefore for some purposes it is reasonable to consider certain subalgebras or reduced algebras of less dimensions instead of working with the whole expanded algebra .
the problem of the construction of algebras of less dimension that are related to an expanded lie algebra is rather complicated .
two classes of such algebras were presented in @xcite for @xmath0-expanded lie algebras admitting an agreed decomposition of the associated semigroups and lie algebras .
a semigroup @xmath0 and a lie algebra @xmath3 are said to admit _ a resonant subset decomposition _
if they can be decomposed as @xmath34 ( where the direct sum is interpreted in the sense of vector spaces only ) and @xmath35 with some index set @xmath36 in such a way that @xmath37\subset\bigoplus_{r\in i(p , q)}v_r$ ] and @xmath38 , where for every @xmath39 the index set @xmath40 is a subset of @xmath36 . in the resonance case , @xmath41 is a _ resonance subalgebra _ of the @xmath0-expanded lie algebra @xmath21 . for each resonant decomposition
this gives a lie algebra of dimension less than the dimension of @xmath21 .
algebras of the other class are constructed by means of the reduction of lie algebras .
suppose that for each @xmath42 the set @xmath43 is partitioned into the subsets @xmath44 and @xmath45 such that @xmath46 .
this partition induces the decomposition @xmath47 , where direct sums are interpreted in the sense of vector spaces , @xmath48 and @xmath49 . as @xmath50\subset\hat{\mathfrak g}}$ ] , the projection of the lie bracket of @xmath51 to @xmath52 gives a well - defined lie bracket on @xmath53 .
in other words , the lie algebra @xmath52 with this bracket is the reduced algebra for @xmath51 . given an abelian semigroup @xmath0 with a zero @xmath54 , i.e. , @xmath55 for all @xmath56 , any @xmath0-expanded lie algebra @xmath57 can be decomposed as @xmath58 with @xmath26\subset \hat{v}$ ] , where @xmath59 . in this case
the reduced lie algebra @xmath60 is called the _
@xmath61-reduced algebra _ of @xmath62 .
[ nesterenko : right_side ] the condition ( [ nesterenko : cijk ] ) implies that the @xmath61-reduced algebra @xmath63 has the following commutation relations in the basis @xmath64 : @xmath65= \left\ { \begin{array}{ll } \sum\limits_{k=1}^nc_{ij}^k \lambda_\alpha\lambda_\beta e_k \quad & \text{if } \quad \lambda_\alpha\lambda_\beta\ne 0_s,\\ 0\quad & \text{if } \quad \lambda_\alpha\lambda_\beta=0_s .
\end{array } \right.\ ] ] as generalized inn
wigner contractions are related to gradings of the contracted lie algebras ( i.e. , to special decompositions of the underlying vector space ) , it is clear that all generalized iw - contractions can be obtained by means of resonant @xmath0-expansions .
see @xcite for details . at the same time , this claim does not imply that @xmath0-expansions exhaust all possible contractions since there exist contractions that are not realized by generalized iw - contractions @xcite .
consider the example of such a contraction constructed in @xcite for the seven - dimensional lie algebras @xmath66 and @xmath67 .
these algebras are defined by the commutation relations @xmath68=e_{i+1 } , \
, 2\leqslant i \leqslant 6,\\ \phantom{\mathfrak g_f\colon } [ e_2,e_3]=e_6 , \quad [ e_2,e_4]=e_7 , \quad [ e_2,e_5]=e_7 , \quad [ e_3,e_4]=-e_7 , \\
\mathfrak g_e\colon [ e_1,e_i]=e_{i+1 } , \ , 2\leqslant i \leqslant 6 , \quad [ e_2,e_3]=e_6+e_7 , \quad [ e_2,e_4]=e_7\end{gathered}\ ] ] and are characteristically nilpotent since their differentiation algebras @xmath69 @xmath70 are nilpotent .
here @xmath71 denotes the @xmath72 matrix with only nonzero @xmath73th entry that equals 1 .
( it is obvious that both the differentiation algebras consist of lower triangular matrices . )
this means that the algebras @xmath66 and @xmath67 admit no grading and hence none of them can be a result of a generalized iw - contraction . on the other hand , there exist the contraction from @xmath66 to @xmath67 provided by the following contraction matrix at @xmath74 : @xmath75 examples on non - universality of the generalized iw - contractions in dimension four can be found in @xcite .
this is why it is still unclear whether all contractions can be obtained via @xmath0-expansions . on the other part
, there exist @xmath0-expansions which are not equivalent to contractions .
see , e.g. , the example on a connection between the lie algebras @xmath76 and @xmath77 that is presented in the next section .
other objects closely related to @xmath0-expansions are given by the purely algebraic notion of _ graded contractions _
the graded contraction procedure is the following .
structure constants of a graded lie algebra are multiplied by numbers which are chosen in such a way that the multiplied structure constants define a lie algebra with the same grading .
graded contractions include discrete contractions as a subcase but do not cover all continuous ones .
contractions of real and complex three - dimensional lie algebras were completely studied in @xcite .
as all these contractions are realized by the generalized inn
wigner contractions , they can be realized by @xmath0-expansions .
moreover , lead to establishing new relations between such algebras .
according to mubarakzyanov s classification @xcite , nonisomorphic real three - dimensional lie algebras are exhausted by the two simple algebras @xmath76 and @xmath78 ; the two parameterized series of solvable algebras @xmath79 and @xmath80 ; the three single indecomposable solvable algebras @xmath81 , @xmath82 and @xmath83 ; and the two decomposable solvable algebras @xmath84 and @xmath77 . in
@xcite it was proven that all unimodular three - dimensional algebras ( namely , @xmath85 , @xmath78 , @xmath86 , @xmath87 , @xmath81 and @xmath84 ) belongs to the orbit closure of at least one of the simple algebras .
all contractions are shown on figure [ nesterenko : fig1 ] by dashed lines ( each arrow indicates the direction of the corresponding contraction ) . repeated contractions ( i.e. , contractions of the kind : @xmath3 contracts to @xmath88 if @xmath3 contracts to @xmath89 and @xmath89 contracts to @xmath88 ) are implied but not indicated on this figure .
all three - dimensional contractions were realized by generalized iw - contractions , therefore the corresponding @xmath0-expansions are also exist .
there are a number of necessary conditions to be satisfied for a pair of lie algebras connected by a contraction , see , e.g. , @xcite .
the following examples show that @xmath0-expansions of lie algebras do not obey the major part of these rules .
these examples involve the algebras ( for each algebra we present only nonzero commutation ralations ) @xmath90=e_1,\ [ e_2,e_3]=e_3,\ [ e_1,e_3]=2e_2 ; \\ \label{nesterenko : a21 } a_{2.1}\oplus a_1\colon\quad [ e_1,e_2]=e_1 ; \\ \label{nesterenko : a33 } a_{3.3}\colon\quad [ e_1,e_3]=e_1,\ [ e_2,e_3]=e_2.\end{gathered}\ ] ] using ( [ nesterenko : tr_ad ] ) for the basis elements @xmath93 of the @xmath0-expanded lie algebra @xmath62 , we get @xmath94 where @xmath95 thereby the unimodularity property is necessarily preserved by any sole @xmath0-expansion . at the same time
, a unimodular lie algebra may contain a non - unimodular subalgebra or may be reduced to a non - unimodular lie algebra .
consider the expansion of the lie algebra @xmath76 by means of the abelian semigroup where @xmath96 , i.e. , @xmath97 is the zero element @xmath98 of the semigroup .
the elements @xmath99 , @xmath100 , @xmath101 , form a basis of six - dimensional @xmath102-expanded lie algebra @xmath62 . then from ( [ nesterenko : sl2r ] ) we obtain the nonzero commutation relations @xmath103=e_2 $ , & $ [ e_1,e_4]=e_2 $ , & $ [ e_1,e_5]=2e_4 $ , & $ [ e_1,e_6]=2e_4$,\\[.7ex ] $ [ e_2,e_3]=e_2 $ , & $ [ e_2,e_4]=e_2 $ , & $ [ e_2,e_5]=2e_4 $ , & $ [ e_2,e_6]=2e_4$,\\[.7ex ] $ [ e_3,e_5]=e_6 $ , & $ [ e_3,e_6]=e_6 $ , & $ [ e_4,e_5]=e_6 $ , & $ [ e_4,e_6]=e_6$. \end{tabular}\ ] ] elements @xmath104 , @xmath105 and @xmath106 span a three - dimensional subalgebra with the nonzero commutation relations @xmath107=e_2 $ ] , @xmath108=e_2 $ ] .
the basis change @xmath109 , @xmath110 , @xmath111 leads to the unique nonzero relation @xmath112=\tilde{e}_1 $ ] and , therefore , we obtain the lie algebra @xmath77 , c.f .
( [ nesterenko : a21 ] ) .
another example concerns the lie algebras @xmath76 and @xmath83 .
let a three - element abelian semigroup with a zero @xmath113 satisfies the conditions @xmath114 and @xmath115 .
the above properties of @xmath116 are consistent with the semigroup structure and are sufficient for defining @xmath116-expansions .
thus , the @xmath116-expansion of @xmath76 is the nine - dimensional lie algebra @xmath117 with the nonzero commutation relations @xmath118=e_1 $ , & $ [ e_1,e_5]=e_1 $ , & $ [ e_1,e_6]=e_1 $ , \\[.7ex ] $ [ e_1,e_7]=2e_4 $ , & $ [ e_1,e_8]=2e_4 $ , & $ [ e_1,e_9]=2e_4 $ , \\[.7ex ] $ [ e_2,e_4]=e_1 $ , & $ [ e_2,e_5]=e_1 $ , & $ [ e_2,e_6]=e_2 $ , \\[.7ex ] $ [ e_2,e_7]=2e_4 $ , & $ [ e_2,e_8]=2e_4 $ , & $ [ e_2,e_9]=2e_5 $ , \\[.7ex ] $ [
e_3,e_4]=e_1 $ , & $ [ e_3,e_5]=e_2 $ , & $ [ e_3,e_6]=e_1 $ , \\[.7ex ] $ [ e_3,e_7]=2e_4 $ , & $ [ e_3,e_8]=2e_5 $ , & $ [ e_3,e_9]=2e_4 $ , \\[.7ex ] $ [ e_4,e_7]=e_7 $ , & $ [ e_4,e_8]=e_7 $ , & $ [ e_4,e_9]=e_7 $ , \\[.7ex ] $ [ e_5,e_7]=e_7 $ , & $ [ e_5,e_8]=e_7 $ , & $ [ e_5,e_9]=e_8 $ , \\[.7ex ] $ [ e_6,e_7]=e_7 $ , & $ [ e_6,e_8]=e_8 $ , & $ [ e_6,e_9]=e_7$. \end{tabular}\ ] ] concerning the rest of non - unimodular lie algebras , namely @xmath82 , @xmath120 and @xmath121 , it seems to be impossible to construct them from simple three - dimensional lie algebras by means of @xmath0-expansion .
this conjecture is motivated by the disagreement of the right - hand sides of the respective canonical commutators that can not be overcome by basis changes .
nevertheless , this conjecture needs a rigorous proof . [
[ s - expansion - from - a_2.1oplus - a_1-to - a_3.3-and - vice - versa ] ] @xmath0-expansion from @xmath77 to @xmath83 and vice versa ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ consider @xmath116-expansion of the lie algebra @xmath77 . to skip the tedious commutation relations of the nine - dimensional lie algebra we consider only those which concern the basis elements @xmath122 , @xmath123 and @xmath124 .
they are @xmath125=e_1,\ ; [ e_2,e_6]=e_2,\ ; [ e_1,e_2]=0.\ ] ] this implies that the basis elements @xmath104 , @xmath105 and @xmath126 form a subalgebra isomorphic to @xmath83 , c.f .
( [ nesterenko : a33 ] ) . the inverse connection can be obtained by means of @xmath102-expansion of the algebra @xmath83 .
the basis elements @xmath104 , @xmath105 and @xmath126 span the subalgebra isomorphic to @xmath77 .
indeed , the nontrivial commutation relations between these elements are @xmath127=0,\quad [ e_1,e_6]=e_1,\quad [ e_2,e_6]=e_1.\ ] ] after the basis change @xmath128 , @xmath129 , @xmath130 , we obtain the unique nonzero commutation relation @xmath112=\tilde{e}_1 $ ] .
in contrast to the contraction procedure , the last expansion creates a lie algebra of more complicated structure .
for example , the dimension of the center decreases from 1 to 0 , the dimension of the cartan subalgebra decreases from 2 to 1 and the dimension of the derivative enlarges from 1 to 2 after the @xmath0-expansion from the lie algebra @xmath77 to the algebra @xmath83 . at the same time , the @xmath0-expansion , even combined with algebra reduction and singling out a subalgebra , preserves certain properties of lie algebras .
thus , we have @xmath131=(ss)\times[\mathfrak g,\mathfrak g]$ ] . hence the algebra
@xmath132 is solvable ( resp .
nilpotent ) if and only if the algebra @xmath3 is , and then the solvability ( resp .
nilpotency ) degrees of these algebras coincide
. the procedures of reducing the algebra @xmath132 and singling out a subalgebra may not increase the solvability ( resp .
nilpotency ) degree .
note that all four of the discussed examples of expansions can be obtained from the non - abelian two - dimensional lie algebra , since in all cases the key role is played by the commutation relation @xmath133=e_1 $ ] .
@xmath0-expansions do not set any ordering relationship on the variety of lie algebras of a fixed dimension .
this statement is illustrated by the example of lie algebras @xmath77 and @xmath83 that can be connected by @xmath0-expansion in both directions .
therefore , in spite of the fact that it is possible to construct non - unimodular lie algebras from unimodular ones , there is still a question whether @xmath0-expansions fit to the classification of solvable lie algebras of a fixed dimension by means of @xmath0-expansions of simple ( semisimple ) lie algebras of the same dimension . | @xmath0-expansions of three - dimensional real lie algebras are considered .
it is shown that the expansion operation allows one to obtain a non - unimodular lie algebra from a unimodular one .
nevertheless @xmath0-expansions define no ordering on the variety of lie algebras of a fixed dimension .
* @xmath0-expansions of three - dimensional lie algebras * * maryna nesterenko * institute of mathematics of nas of ukraine , 3 tereshchenkivska str . , kyiv-4 , 01601 ukraine + e - mail : [email protected] |
the question of how the adsorption of foreign particles affects the properties of materials and the means to control this is of central importance in domains ranging from separation processes to nanotechnology .
this motivates the continuing investigation on the factors determining the adsorption process and the search of conditions most favorable for its control ( composition of the adsorbing fluid , adsorption geometry etc . )
the purpose of this communication is to propose a novel method that allows this , on the basis of monte carlo simulations of a fluid - slit pore equilibrium . to avoid having to consider specific interactions as in molecular adsorption
, we choose to point out the basic mechanisms on a simple model with only hard - sphere and dipolar interactions .
the situation closest to this model is then the adsorption of macroparticles .
another reason is the recent development of studies of colloidal adsorption @xcite .
indeed , while numerous studies exist on molecular adsorption ( see for e.g. @xcite and refs @xcite for more recent work ) , one practical advantage of using colloids is the possibility to tune their effective interaction ( eg . by adding polymeric depletants ) and their coupling with external fields , possibly in confined geometry @xcite .
the actual behavior may , however , be complicated by the interplay of different effects ( see for example the role of static and hydrodynamic forces in microfluidics devices @xcite ) .
we thus propose here a method that allows a fine control of the composition of the adsorbed fluid , while remaining very simple .
we start with the simplest confinement geometry : an open slit pore with parallel walls in equilibrium with a bulk fluid .
it has been used in several theoretical studies to determine the parameters affecting the behavior of the confined fluid , ( see for example gubbins , sarkisov , duda1,jcpabd , jpcgub , virgiliis , duda2,kim and references therein ) . since we seek a method in which the pore geometry , the interactions ( between the particles and the particles and the confining medium ) as well as the bulk thermodynamic state are fixed , one alternative is the coupling with an external field .
this should always be possible since besides particles having a permanent dipole such as magnetic colloids , colloidal particles are always polarizable to some extent .
we thus took a uniform electric field @xmath0 normal to the walls .
as in jcpcharles we considered a mixture in which one species bears a dipole moment @xmath1 , taken permanent for simplicity .
the field is then not applied in the bulk .
we thus have pure hard - spheres ( species 1 ) and dipolar hard - spheres ( species 2 ) , possibly with a non - additive diameter@xmath2 in the potential @xmath3 ( see also the discussion of figure 5 ) .
both species have a hard - sphere interactions with the walls .
this makes the model more appropriate @xcite to a mixture of hard - sphere - like colloidal particles , than to a molecular mixture ( see however the final remarks ) .
the effect of an external field ( and temperature ) on the filling of a cylindrical pore was also studied in @xcite ( see also and bratko for a slit pore ) , but not from a bulk mixture .
previous studies considered the role of the pressure in one - component fluids ( eg . sarkisov ) , or the total density and the mole fractions duda1,duda2,kim in bulk mixtures but without field .
as shown below , the combination of both will play here a crucial role . an inhomogeneous multicomponent mixture with anisotropic interactions being difficult to study by analytical methods ( we are aware of one study by density functional theory of the adsorption from a mixture of polar molecules @xcite ) , we used monte carlo simulation ( see also , for e.g. rasaiah , klapp , rasaiah2,weis , jcpcharles ) .
we already pointed out how the structure can be modulated by the combination of various interactions jcpabd , jpcgub and by the action of the field @xcite .
however , only the density profile of the particles through the pore could be modulated in @xcite since the total number of particles was kept fixed ( simulations in the canonical ensemble ) .
an important difference here is that we consider an open pore which exchanges particles with a reservoir .
one may then achieve much stronger variations of the density of each component in the pore .
the physical pore is assumed large enough that the interfacial region in which it is in contact with the reservoir plays a negligible role @xcite .
for this reason we will refer to the fluid in the reservoir far from this region as the `` bulk '' .
the pore / fluid equilibrium is determined by the equality of the chemical potentials @xmath4 and @xmath5 of both species in the bulk and in the pore .
but since the practical control variables are the total density @xmath6 and mole fraction @xmath7 of the dipolar species , in the bulk , the latter is studied in the canonical ensemble . by considering only homogeneous states or metastable states very close to the coexistence boundary , @xmath8 and @xmath9 in the bulk
are determined with sufficient accuracy from widom s insertion method ( see for eg @xcite for this point ) . @xmath8 and @xmath9 are then used to study the fluid in the pore in the grand - canonical ensemble .
we can then compute the average density of each species in the pore as a function of @xmath6 and @xmath7 .
hereafter , reduced variables @xmath10 and @xmath11 will be used .
the reduced density in the pore is @xmath12 , with @xmath13 the average number of particles for a lateral surface @xmath14 with periodic conditions in the @xmath15 and @xmath16 directions and @xmath17 the accessible volume in the pore .
we took a pore width @xmath18 . in the bulk
, @xmath19 is fixed .
we show in figure 1 the first basic mechanism : a field induced filling of the pore by a one - component dipolar fluid . at increasing field strength @xmath20 ,
the pore is progressively filled by the dipoles with a rate that depends on @xmath21 . for the value of the @xmath22 and @xmath23 used here
, the explanation seems that the field - dipole interaction energy @xmath24 offsets the entropy loss due to their orientation in the direction of the field .
this is more visible at low @xmath25 in which case the slope is nearly constant beyond @xmath26 . for a particle diameter of @xmath27 and @xmath28 , for example
, this corresponds to @xmath29 and @xmath30 .
thanks to the scaling factor @xmath31 in its definition , the same value of @xmath32 is also appropriate for the dipolar interaction between _ molecular _ species . is the total density in the pore and @xmath33 the field strength in reduced units .
the bulk density is @xmath34 ( _ filled squares _ ) and @xmath35 ( _ empty squares _ ) .
the lines are a guide to the eye . ]
the dipole moment being then of the order of one debye , a field strength of the order @xmath36 is needed to obtain the same reduced energy @xmath37 .
such field strengths are not unusual for confinement at the molecular scale ( for water in nanopores see for example @xcite and @xcite , in particular figure 1 in the last one ) .
note that the equilibrium state in presence of the field may not always be the filled one at other parameters or if more complex interactions are considered ( see ref @xcite for example ) , due to the competition between energetic and entropic contributions .
figure 2 illustrates the second mechanism : the relative population of a pore in equilibrium with a mixture having a natural tendency to demix .
the simplest one is the mixture of non - additive hard spheres @xcite in which the cross diameter is @xmath38 previous studies @xcite have shown that when the pore is in equilibrium with a mixture in which one species is in minority ( say @xmath39 for a non additivity parameter @xmath40 ) a population inversion occurs in the pore when the total bulk density is varied .
this occurs here for @xmath41 between @xmath42 and @xmath43 for a pure non - additive hs mixture and between @xmath44 and @xmath42 for the hard - sphere dipole mixture . having the basic ingredients
, we may now combine them to produce the desired effect : by choosing the composition of the bulk fluid so as to be close to the population inversion in the pore , we anticipate that the closer we are from the threshold density the weaker will be the external field @xmath45 required to trigger it .
this is shown in figure 3 . in the most favorable case
shown the actual value of @xmath46 is about @xmath47 .
and bulk densities ( from right to left ) @xmath48 .
the inset shows the corresponding dipoles and hard - spheres density in the pore for @xmath49 . ]
a small variation of the applied field produces the inversion : the dipolar particles are selectively absorbed by a weak change about @xmath46 , the converse being possible , perhaps with some hysteresis @xcite . near the adorption jump ( @xmath50 , @xmath51 for @xmath52 ) a slight change in temperature ( @xmath53 in figure 4 ) produces a detectable change in adsorption . ,
@xmath54 and @xmath55 from left to right .
the bulk mole fraction is @xmath56 and bulk densities are @xmath57 . here
@xmath58 . ]
this observation may be important for some applications ( since the reduced variables combine @xmath59 and @xmath60 with t , while @xmath61 in the bulk , one has to rescale @xmath62 by the factor @xmath63 to compare two temperatures at a given values of @xmath59 ) . finally ,
in order to check the effect of a different interaction @xmath64 between species 1 and 2 , we also show in figure 5 the result for a yukawa repulsive potential .
is the density of the dipolar species , in the pore . _
empty squares _ : one - component dipolar fluid ( for @xmath65 as in figure 1 ) ; _ open circles _ : additive mixture of hard spheres and dipoles ( with @xmath56 and @xmath66 in the bulk ) .
_ diamonds _ : same for a yukawa repulsion between the dipoles and the hard - spheres ; _ filled circles _ : same for the non - additive hard - spheres - dipolar hard - spheres mixture .
the range of the yukawa potential ( with @xmath67 ) gives the same contribution to the second virial coefficient as the non - additive hard - spheres potential with @xmath40 .
_ filled squares _ : one - component dipolar fluid with @xmath66 ( as in figure 1 ) .
] we observe that the phenomenon is quite general .
the requirement for observing a sensitive field effect is that the self coordination should be more favored in the mixture .
the poor miscibility can be favored by suitable chemical composition of the particles surface layers pusey , hennequin .
weak specific interaction with the pore walls can be achieved similarly . as our main goal was
to demonstrate the phenomenon of a field activated adsorption from an unstable mixture , we used a simple simulation strategy .
accordingly , we did not conduct a detailed study of the behavior of the confined fluid .
for instance , phase equilibrium in the pore may take place before the spontaneous condensation ( adsorption jump ) predicted in the grand canonical simulation @xcite . according to duda al.@xcite ,
the inversion line for non - additive hard spheres is close to the bulk fluid coexistence line but the two phenomena are different .
we actually observed that the inversion corresponds to a bulk fluid close to demixing or slightly in the two - phase region .
regardless of this , the essential point is that the density in the liquid - like phase should be close to the value after the adsorption jump , as in one - component systems @xcite .
the precise relation between these observations and other phenomena such as capillary condensation , wetting , hysteresis , etc .
( see e.g. wetting , neimarkpre , kierlik ) will be discussed in future work .
in conclusion , these results show that the combination of two generic mechanisms allows a quite sensitive control of the pore filling .
although this method has been demonstrated for particles that are closest to the optimum conditions ( ie hard - sphere - like colloids ) , none of them is exclusive and since the basic mechanisms ( demixing instability and coupling with an external field ) are quite generic , this prediction should concern a broader class of systems ( including molecular ones ) . in order to benefit from the field effect , one species should be either polar ( e.g. ferrocolloid in magnetic fields ) or much more polarizable than the other ( the results given here being relative to permanent dipoles ) .
the solution should also not contain free charges to avoid particle motion due to the action of the field ( electrohydrodynamic flows ) , not considered in this simple model .
polar molecules being on the other hand rather common , one should consider in this case also the role of specific interactions .
we believe that further experimental studies and simulations on this method are worthwhile given the diversity of possible applications of this field controlled composition of the confined fluid and hence flexible control of the physical properties that depend on the composition of the confined fluid .
just as an example , one may consider to modulate in this way the dielectric response of the confined fluid for optical applications .
f. k. katsaros , th .
a. steriotis , g. e. romanos , m. konstantakou , a. k. stubos , and n. k. kanellopoulos , microporous and mesoporous materials * 99 * , 181 ( 2007 ) ; j. pires , m. bestilleiro , m.pinto , and a. gil , separation and purification technology * 61 * , 16 ( 2008 ) ; a. busch , y. gensterblum , b. m. krooss , and n. siemons , int .
j. of coal geology * 66 * , 53 ( 2006 ) . | starting from a generic model of a pore / bulk mixture equilibrium , we propose a novel method for modulating the composition of the confined fluid without having to modify the bulk state . to achieve this , two basic mechanisms - sensitivity of the pore filling to the bulk thermodynamic state and electric field effect -
are combined .
we show by monte carlo simulation that the composition can be controlled both in a continuous and in a jumpwise way . near the bulk demixing instability ,
we demonstrate a field induced population inversion in the pore .
the conditions for the realization of this method should be best met with colloids , but being based on robust and generic mechanisms , it should also be applicable to some molecular fluids . |
ground - based long - slit spectroscopy of the extended narrow line region ( enlr a few kpc scale ) of the seyfert 1 galaxy ngc4151 has shown that the kinematics , ionization structure and morphology of the emission - line gas are consistent with photoionization of gas in the galactic disk by an anisotropic radiation field produced by the active nucleus ( penston et al .
the situation on the scale of the narrow line region ( sub - kpc ) is however , much less clear as there is a large misalignement between the enlr at pa @xmath0 50 ( prez et al .
1989 ; robinson et al . 1994 ) and the nlr which is co - spatial with the radio jet at pa @xmath0 77 ( pedlar et al .
if both the nlr and enlr structures are to be explained in terms of an `` ionization cone '' then this misalignment implies either a very broad ( opening angle @xmath0 120 ) cone of ionizing radiation , intercepting the disk at grazing incidence ( pedlar et al .
1992 , vila - vilar 1993 , robinson et al . 1994 ) or a geometry which is distinct from the `` standard '' unified model which has the cone axis fixed to that of the radio ejecta ( evans et al .
1993 ) .
the complex structure of clouds and filaments in the inner 5 of the nlr revealed by _ hubble space telescope _ ( hst ) images suggests that interaction with the radio jet may have an important influence on the observed morphology . in this letter , we present the first results of new hst long - slit spectroscopy of the central 8 of ngc4151 at a spatial resolution of 0.029 which enables us to separate for the first time the emission from the individual nlr clouds , and demonstrates that the interaction with the radio ejecta strongly affects the local kinematic and ionization conditions of the emission gas .
our results imply that it is this interaction , rather than anisotropic illumination , which determines the basic structure of the nlr . for a distance to ngc4151 of 13.3 mpc
, 0.1 corresponds to a linear scale of 6.4 pc in the plane of the sky .
the spectroscopic data consist of four spectra obtained with the faint object camera ( foc ) f/48 long - slit facility on board the hst .
the slit width is 0.063 , and its useful length about 10 .
we have also used foc f/96 images with filters centered at the [ ] @xmath25007 emission - line and nearby continuum .
table [ log_tab ] summarizes the relevant information on the above data set .
the f/96 images were reduced using the standard pipeline procedure , as described in the foc handbook ( nota et al .
1996 ) .
the spectra were obtained with the slit at pa=47 , and cover the interval 3650 5470 .
the data reduction followed the same procedure as described in macchetto et al .
( 1997 ) ; wavelength resolution is 5.1 @xmath0 300 kms@xmath1 at [ ] @xmath25007 , and the spatial psf is of the order of 0.08 ; flux calibration was obtained using the iue standard lds749b . by comparing the brightness profile of the [ ] @xmath25007 emission line on the spectra with that on the f/96 image
, we were able to pinpoint the location of the slit to within 0.06 ( 2 pixels at the f/48 resolution ) .
figure [ plate1_fig ] ( plate [ plate1_fig ] ) shows the slit positions superimposed on the [ ] @xmath25007 image .
the spectra are located on the nucleus , and 0.23 , 0.40 and 0.46 nw of it , and will be hereafter referred to as pos1 , pos2 , pos3 and pos4 , respectively .
the nuclear spectrum was heavily saturated in the central 15 pixels , due to the strong continuum emission .
the presence of a complex radial velocity structure is evident in all the brightest narrow emission lines , such as [ ] , [ ] , and [ ] . to study the detailed kinematics of the gas the [ ] @xmath24959,5007 emission lines , which provide the best resolution and s / n ratio , were extracted in spatial windows varying from two to 20 pixels along the slit , and both lines fitted simultaneously using two to six gaussian components , constraining the physical parameters of the corresponding components of each line to their theoretical values ( 3:1 intensity ratio , 48 separation , and same fwhm ) .
figure [ plate2_fig ] ( plate [ plate2_fig ] ) shows the combined vla@xmath4merlin 5 ghz radio map from pedlar et al .
( 1993 ) , overlaid on the [ ] @xmath25007 image of ngc4151 .
the registration was made assuming that the optical nucleus corresponds to component c4 , as advocated by mundell et al .
there is an apparent anti - correlation between the optical and the radio emission , with the brightest emission - line filaments surrounding the radio knots as would be expected if the radio - jet is clearing a channel in the surrounding medium and enhancing the line emission along its edges by compression of the ambient gas .
figure [ rotcurv_fig ] shows the results of the [ ] @xmath24959,5007 fitting for three of the four slit positions ( the pos3 and pos4 data are very similar to each other and do not give independent information ) .
the rotation curves were constructed assuming that the narrowest , closest to the systemic velocity , and usually the brightest component ( filled squares ) represent the `` main '' nlr gas emission .
the curves are plotted as a function to the distance to the center of the slit , defined as a line passing through the nucleus , perpendicular to the slit direction . the solid line on each panel represents the integrated [ ] @xmath25007 spatial profile .
the general behavior of the main component is consistent with normal galactic rotation as defined by the enlr gas emission ( robinson et al .
1994 ) , with strong but localized perturbations superimposed , represented by one , or sometimes two , narrow ( fwhm @xmath5 800 kms@xmath1 ) secondary components either blue- or redshifted relative to the core of the line ( shown as open circles and squares in figure [ rotcurv_fig ] ) . in the inner 2 of the nuclear spectrum
, the behaviour of the emission gas shows a strong deviation from the outermost rotation curve .
it is important to note that the most kinematically disturbed features are found to be associated with the emission clouds located along the edges of the strongest radio emission components : the red / blue - shifted knots at @xmath0 1 sw of the center of the slit in the pos2 spectrum are located at one side of the c2 radio component , and are well fitted by two gaussian components centered at @xmath41300 and @xmath6800 kms@xmath1 from the core velocity , and fwhm @xmath0 330 and 490 kms@xmath1 , respectively ; the high velocity ( @xmath7v @xmath0 @xmath61700 kms@xmath1 ) cloud * ( b ) * in pos3 lies also alongside this same radio component ; the second - brightest radio component c3 is in front of the bright emission complex located 0.20.5 nw of the nucleus , where we observe a clear double - peaked profile in all emission lines , well represented by two narrow ( fwhm @xmath0 430kms@xmath1 ) components , one of which is blueshifted by @xmath0 660 kms@xmath1 from the core velocity .
all these features are characterized by a very small projected size : the two high - velocity clouds in pos2 discussed above are 0.17 to 0.23 ( less than 15 pc ) across , while the double - peaked feature in pos3 ( see figure [ sampspec_fig]c ) is resolved along a 0.3 ( 20 pc ) bin , giving further support to the idea that the nlr morphology is strongly shaped in walls and voids by the expanding , fast - moving radio components .
if the localized kinematics perturbations described above are a consequence of fast shocks created by interactions with ejected radio material the physical conditions in the clouds may also be influenced by the shocks .
not only can the shocks compression lead to local density enhancements ( taylor , dyson & axon 1992 ) but these may also be accompanied by secondary photoionizing radiation ( sutherland , bicknell & dopita 1993 ; morse , raymond & wilson 1996 ) . we clearly see that the emission - line ratios vary substantially on scales of a few tenths of an arcsec , indicating that the density and ionization state of the emission gas are strongly influenced by the local conditions .
we also detect several examples of off - nuclear blue - continuum emission , consistently associated with the brightest emission - line knots .
figure [ sampspec_fig ] presents a sample of the spectra observed , corresponding to the regions * ( a ) * to * ( c ) * in figure [ rotcurv_fig ] .
the observed line intensities , relative to h@xmath3 = 1.0 , are given in table [ flux_tab ] .
as a reference we included the `` average '' nlr line intensities , obtained by adding up the spatially integrated emission of the four slit positions , and the ground - based data for the enlr ( @xmath8 ) from penston et al .
errors are of the order of 0.05 or smaller for the line ratios , and 0.002 or smaller for the [ ] @xmath24363/[]@xmath25007@xmath44959 ratio .
spectrum [ sampspec_fig](a ) corresponds to the brightest [ ] @xmath25007 knot in pos2 , one of the clouds associated with the radio component c3 , located at a projected distance of 0.18 sw ( 12 pc ) of the nucleus .
spectrum [ sampspec_fig](b ) corresponds to the region with the high velocity cloud located 1 sw ( 64 pc ) of the center of pos3 .
this component can be clearly seen in the [ ] profile .
the fwhm of the `` main '' [ ] @xmath25007 emission in this spectrum is unresolved ( 310 kms@xmath1 ) , while spectra from neighboring regions ( 6 pixels , 0.17 @xmath0 11 pc either side ) show a @xmath25007 line twice as broad .
the h@xmath9/h@xmath3 ratios in these two positions are consistent with little or no reddening .
all line ratios for spectrum [ sampspec_fig](b ) are for the core component only , and do not include the high velocity cloud contribution . both of these spectra present similar ratios [ ] @xmath23869/h@xmath3 , while [ ] /[]@xmath23727 in * ( a ) * is about twice that in * ( b)*. the high value of the @xmath23869/@xmath23727 ratio is a local characteristic of region * ( a ) * that is not observed in nearby clouds , and there is no clear evidence of kinematic perturbations in this spectrum .
the fact that @xmath24686 in * ( a ) * is not significantly stronger than in * ( b ) * points against the hypothesis of a matter - bounded cloud , which will enhance the [ ] emission relative to low - ionization lines ( binette , wilson & storchi - bergmann 1996 ) .
it is also possible that the interaction with the radio - jet would compress the gas , and thus lower the ionizaton parameter , favoring the emission of low - ionization lines such as [ ] in region * ( b)*. notice , however , that the @xmath23869/@xmath23727 ratio for this region is similar to the ratio averaged over all the observed area , where one would expect the effects of local density enhancements to have been washed out . for these two spectra
we also measured ratios @xmath10 } \equiv $ ] [ ] @xmath24363/[]@xmath25007@xmath44959 @xmath0 0.012 and 0.016 , respectively . using the 3-level atom of mccall ( 1984 ) , for a reasonable electron density range of @xmath11 to @xmath12 @xmath13 these values correspond to [ ] temperatures @xmath14}$ ] in the range 12,000 to 16,000 k , slightly above the limit of @xmath15 11,000 k predicted by most photoionization models , but still compatible with the value found for the purportedly photoionized , kinematically undisturbed enlr .
in fact , the overall similarity of the nlr line ratios to those of the enlr suggests that variations in electron denstities created by different degrees in shock compression in individual clouds are probably the main cause of the observed variations of ionization from knot to knot .
however , this is not the whole story as shown in spectrum [ sampspec_fig](c ) which , in addition to double - peaked profiles , shows clear evidence of a localized blue continuum . such localized regions of off - nuclear blue continuum emission are seen in several positions in our spectra , and are always associated to , but not exactly coincident with , an emission - line knot . as we mentioned above ,
this is a characteristic spectral signature of the free - free emission of the hot gas created by the fast shocks and therefore provides concrete evidence for the existence of a local ionizing continuum which might be playing a significant role in the ionization structure of the individual clouds .
the high spatial resolution long - slit spectroscopy of the nuclear region of ngc4151 presented in this paper clearly show that the popular picture of anisotropic illumination by the nuclear source is overly simplistic to explain the complex morphology and kinematics observed in the nlr .
comparison of the radio and optical data shows that the line emission is enhanced along the edges of the radio knots , as would be expected to result from the interaction between the jet and the surrounding medium , similar to what is observed for ngc1068 ( gallimore et al . 1996 ; capetti , macchetto & lattanzi 1997 ; capetti , axon & macchetto 1997 ) .
the general behaviour of the [ ] @xmath25007 velocity field is consistent with disk rotation , with local deviations caused by turbulence ( probably associated with the sweeping / expanding motion of the radio cocoon ) .
we detect several distinct and very localized high velocity ( @xmath61700 to @xmath41300 kms@xmath1 ) , narrow ( fwhm @xmath5 600 kms@xmath1 ) emission components , both blue and redshifted , confined in projected dimensions of up to 0.3 .
we find no evidence for systematic red / blue - shifted deviations as expected by the presence of a generalized outflow , as advocated by schulz ( 1990 ) .
detailed modeling of the inner @xmath16 kinematics will be presented in a forthcoming paper .
the data shows that the emission - line ratios can vary substantially within a few tenths of arcsec , indicating that the physical conditions of the emission gas are also strongly influenced by the local conditions .
furthermore , we detect off - nuclear blue continuum emission , consistently associated with the brightest emission - line knots .
the clear morphological association between the radio and the optical emission , and the highly disturbed kinematical components in the nlr of ngc4151 revealed by our data imply that interaction with the radio plasma rather than illumination shapes the nlr . within the context of the present data , the relative balance between the two possible explanations of the ionization structure variations we have discussed , density fluctuations and/or local ionization , can not be resolved as we do not have density measurements of the individual clouds .
nevertheless , it is evident that shocks associated with the jet - cloud interactions play an important role in determining the physical conditions of the _ individual _ emission clouds ( taylor , dyson & axon 1992 ; sutherland at al .
1993 ) . in order to unambiguously determine
if local ionization effects play a significant role in modifying the ionization structure of individual clouds high quality measurements of the electron density in the clouds need to be obtained from future hst spectroscopy .
binnete , l. , wilson , a. s. , & storchi - bergmann , t. 1996 , , 312 , 365 .
capetti , a. , axon , d. j. , & macchetto , f. d. 1997 , , in press .
capetti , a. , macchetto , f.d . , & lattanzi , m. g. 1997 , , 476 , l67 .
evans , i. n. , tsvetanov , z. , kriss , g. a. , ford , h. c. , caganoff , s. , & koratkar , a. p. 1993
, , 417 , 82 .
gallimore , j. f. , baum , s. a. , odea , c. p. , & pedlar , a. 1996 , , 458 , 136 .
macchetto , f. d. , marconi , a. , axon , d. j. , capetti , a. , sparks , w. b. , & crane , p. 1997
, , in press .
mccall , m. l. 1984 , , 208 , 253 .
morse , j. a. , raymond , j. c. , & wilson , a. s. 1996 , , 108 , 426 .
mundell , c. g. , peddlar , a. , baum , s. , odea , c. p. , gallimore , j. f. , & brinks , e. 1995 , , 272 , 355 .
nota , a. et al .
1996 , foc instrument handbook , version 7.0 ( baltimore : stsci ) .
pedlar , a. , howley , p. , axon , d. j. , & unger , s. w. 1992 , , 259 , 369 .
pedlar , a. , kukula , m. j. , longley , d. p. t. , muxlow , t. w. b. , axon , d. j. , baum , s. , odea , c. , & unger , s. w. 1993 , , 263 , 471 .
penston , m.v .
et al . 1990 , , 236 , 53 .
prez , e. , gonzlez - delgado , r. , tadhunter , c. , & tsvetanov , z. 1989 , , 241 , 31p .
robinson , a. et al .
1994 , , 291 , 351 .
schulz , h. 1990 , , 99 , 1442 .
sutherland , r. s. , bicknell , g. v. , & dopita , m. a. 1993 , , 414 , 510 .
taylor , d. , dyson , j. e. , & axon , d. j. 1992 , , 255 , 351 .
vila - vilar , b. 1993 , ph.d .
thesis , univ .
la laguna , tenerife .
lcccr spectra & 1996 jul 03 & x38i0102 t & f305lp & 1247 & & x38i0108 t & f305lp & 697 & & x38i0109 t & f305lp & 697 & & x38i010at & f305lp & 697 images & 1994 dec 03 & x2740101 t & f501n & 2160 & & x2740102 t & f4nd@xmath4f501n & 900 & & x2740108 t & f550 m & 1340 lccccccc [ ] @xmath23727 & 1.54 & 3.33 & 2.09 & 1.21 & 1.79 & 2.56 & 2.44[]@xmath23869 & 1.24 & 1.25 & 0.91 & 0.82 & 0.88 & 1.07 & 1.0h@xmath9 & 0.44 & 0.53 & 0.39 & 0.34 & 0.37 & 0.39 & 0.42[]@xmath24363 & 0.19 & 0.34 & 0.37 & 0.27 & 0.27 & 0.25 & 0.24@xmath24686 & 0.14 & 0.19 & 0.26 & 0.31 & 0.28 & 0.27 & 0.38[]@xmath25007 & 11.9 & 10.8 & 15.2 & 10.8 & 13.7 & 12.5 & 12.9 @xmath23727/@xmath25007 & 0.13 & 0.31 & 0.14 & 0.11 & 0.13 & 0.21 & 0.19@xmath23869/@xmath23727 & 0.81 & 0.38 & 0.44 & 0.68 & 0.49 & 0.42 & 0.41 @xmath10}$ ] & 0.012 & 0.016 & 0.014 & 0.018 & 0.015 & 0.015 & 0.014 | we present the first results of the hubble space telescope / faint object camera long - slit spectroscopy of the inner 8 of the narrow line region of ngc4151 at a spatial resolution of 0.029 .
the emission gas is characterized by an underlying general orderly behaviour , consistent with galactic rotation , over which are superposed kinematically distinct and strongly localized emission structures .
high velocity components shifted up to @xmath0 1500 kms@xmath1 from the systemic velocity are seen , associated with individual clouds located preferentially along the edges of the radio knots .
off - nuclear blue continuum emission is also observed , associated with the brightest emission line clouds .
emission line ratios like [ ] @xmath23869/[]@xmath23727 , and [ ] @xmath23727/h@xmath3 are observed to vary substantially between individual clouds .
we advance the general picture that , as in other seyfert galaxies observed with hst ( e.g. ngc1068 , mrk573 ) , the interaction of the radio jet with the ambient gas strongly influences both the morphology and the physical conditions of the nlr . |
future high - precision measurements of the fluctuations in the cosmic microwave background ( cmb ) may allow key cosmological parameters ( such as @xmath7 , @xmath8 , @xmath9 , the hubble constant , ) to be measured with unprecedented accuracy ( jungman 1996 ; bond 1997 ; zaldarriaga 1997 ) , and a number of ground- , balloon- and space - based missions are currently being planned for this purpose .
both when designing such missions and when analyzing the data sets that they produce , it is important that all relevant sources of foreground contamination are well understood , to minimize the risk that foreground signals are misinterpreted as cmb fluctuations .
since numerous experiments are are currently in the planning and design stages , it is therefore timely to catalog and quantify all possible foreground effects .
summarizes recent foreground estimates from tegmark & efstathiou ( 1996 ) , hereafter te96 " , and bersanelli ( 1996 ) .
the purpose of this _ letter _ is to investigate yet another foreground effect which has not been previously studied : that of weak gravitational lensing of point sources
. in other words , we will discuss the extent to which the point source region ( which occupies mainly the lower right corner of ) is expanded by lensing . since it is impossible to remove all point sources ( as there are for all practical purposes infinitely many , and not all with the same spectra ) , the standard procedure in cmb experiments is to remove all point sources below some flux cut @xmath0 , either by subtracting their estimated emission or by discarding all contaminated pixels .
the total sky brightness from the remaining sources tends to be dominated by faint ones , whereas the brightness fluctuations ( which contribute to ) are dominated by the sources just below the flux cut . for this reason ,
the average brightness @xmath10 exceeds the fluctuations by a substantial factor , typically @xmath11 , as shown in .
a process that caused even minor spatial modulations of @xmath10 could therefore pose a serious problem for cmb experiments .
weak gravitational lensing by large - scale structure ( see villumsen 1996 for a recent discussion ) can affect observed power spectra in more ways than one . if we imagine a pattern painted on the inside of a rubber ( celestial ) sphere , weak lensing corresponds to stretching and compressing the rubber in a random fashion , much like the way our mirror images get distorted by non - flat mirrors in amusement parks .
for weak lensing , this distortion will always constitute a one - to - one mapping of the image , , there will be no caustics where the image folds over " on itself .
there is thus no smearing involved , so the fluctuation power is conserved .
instead , there is a robin hood effect " where power is redistributed between multipoles , from those which have more to those which have less ( seljak 1996 ) . for the cmb power spectrum ,
this effect is typically a few percent , and may well be detectable in future cmb experiments . for point sources , on the other hand , this effect is completely absent : since they have a poisson power spectrum with @xmath12 constant , the robin hood effect ( which effectively simply smoothes the power spectrum ) , will leave @xmath12 unaffected .
however , the presence of a flux cut for point sources produce a different effect , which is absent for the cmb . in those regions where lensing causes magnification , we will see ( and remove ) disproportionately many sources above the flux cut ( turner 1984 , broadhurst 1995 ) .
since lensing leaves the average brightness unaffected , the sky brightness from unremoved sources will be lower than average in this region .
in other words , weak lensing causes the type of modulation of @xmath10 that we warned about above .
the rest of this _ letter _ is organized as follows .
we present a detailed calculation of this effect in section 2 , assess its importance in section 3 and summarize our findings in section 4 .
let the _ unlensed source count function _
@xmath13 denote the number of sources per steradian whose flux would exceed @xmath14 in the absence of lensing .
when weak gravitational lensing magnifies a patch of sky with a magnification factor @xmath15 , this has two separate effects on @xmath13 : * the sources become a factor @xmath15 brighter .
* the sources are seen further apart , reducing their number density by a factor @xmath15 . in summary , the _ lensed source count function _ ,
which we denote @xmath16 , is given by ( > ) = 1 ( > ) . defining the _
differential source count _ as @xmath17 ( a positive function ) , we thus have ( ) = 1 ^ 2 ( ) . for weak lensing , we can write @xmath18 , where @xmath19 ( broadhurst 1995 ) , and we will make the approximation of dropping all quadratic and higher order terms in @xmath20 . using @xmath21 , we can rewrite as ( , ) = ( ) - 1 ( ) , where we have explicitly indicated the fact that the magnification field depends on @xmath22 , the unit vector pointing in our direction of observation .
since the total sky brightness @xmath23 is finite , the quantity @xmath24 must clearly approach zero both as @xmath25 and as @xmath26 , otherwise the integral would diverge at the faint or bright end . since the second term in becomes a total differential when multiplied by @xmath14 , this means that the total brightness is not affected by lensing .
this well - known fact is also verified by integrating @xmath27 times and changing variables . however , when the upper integration limit is not @xmath28 , the same procedure shows that the brightness contribution from all sources below some fixed flux cut _ is _ affected by lensing , and we will now evaluate this effect . for cmb purposes
, we can to a good approximation ( te96 ) model the locations of extragalactic point sources as completely uncorrelated . in the absence of lensing , we can therefore write the observed density of point sources above some flux cut @xmath0 as a sum of angular delta functions , @xmath29 , and model @xmath30 as a poisson process . a poisson process satisfies ( see appendix a of feldman , kaiser & peacock 1994 ) _ p&= & ( ) , + [ poissoneq2 ] _ p&=&()()+ ( , ) ( ) .
we use @xmath31 to denote ensemble averages with respect to the poisson process , to distinguish these from ensemble averages with respect to the random field @xmath20 , which we will denote by @xmath32 .
when performing both averages below , we will omit subscripts and write @xmath33 .
magnification bias forces us to take into account that the total source population is the union of a number of independent subpopulations , corresponding to different flux classes .
we therefore write n ( ) = _ 0^n(,)d , where @xmath34 is a sum of delta functions corresponding to the sources whose fluxes fall between @xmath14 and @xmath35 . with this notation ,
equations and become generalized to _ p&=& ( , ) , + [ poissoneq4 ] _ p&=&(,)(,)+ ( -)(,) ( , )
. the observed sky brightness @xmath36 in a direction @xmath22 is clearly given by ( ) _ 0^n(,)d , and we will now calculate its statistical properties .
using equations , and , we see that the mean is given by = _
0^_f d = - _ f , where we have defined ( ) ()^2 ( ) and where @xmath10 denotes the average total brightness due to sources below our flux cut , , b_0^()d . since @xmath37 ( there is just as much positive as negative weak lensing from large - scale structure ) , the second term in vanishes , and we see that regardless of the flux cut , lensing has no impact on the average brightness of unremoved point sources .
we now turn to the correlation function . using @xmath37 and equations , , and , we obtain & = & _
0^_0^_f dd + & = & _ f + ( , ) , where _ 0^()^2 dis the familiar power spectrum of unresolved point sources in the absence of lensing that was derived in te96 . using @xmath38 again , this reduces to = ^2 + + ()^2 ^ 4_f .
since the first term is merely the familiar and uninteresting monopole , the new effect that we have computed corresponds to the third term .
thus the effect of lensing is to add to the original point source correlation function a new term which is the magnification correlation times @xmath39 .
let us expand the sky brightness @xmath36 in spherical harmonics and investigate the statistical properties of the expansion coefficients ^ * ( ) ( ) d. assuming that the statistical properties of the magnification field @xmath20 are isotropic ( there is no need to assume that @xmath20 is gaussian ) , its corresponding expansion coefficients @xmath40 must satisfy _ f = _ _mmfor some @xmath41 which we will refer to as the _ magnification power spectrum_. substituting into = ^*()( ) dd , we obtain = _ _mm , where the point source power spectrum is = 4_0_0 m b^2 + + ()^2 ^ 4 .
the first two terms are merely the monopole and the shot noise power ( which is independent of @xmath42 ) , derived in te96 , so the new effect due to lensing is given by the third term .
to clarify the relative importance of the last two terms in , it is convenient to express them in terms of quantities that are more directly linked to observations . for typical scenarios , the differential source count @xmath43 is a smooth function on a log - log - plot , which means that near the flux cut @xmath0 , we can approximate it with a power law ( ) ()()^- for some constant @xmath44 . to avoid the above - mentioned divergence of the total brightness
, we must have a logarithmic slope @xmath45 at the faint end and @xmath46 at the bright end .
for instance , for radio sources at 1.5 ghz , the vla first survey gives the logarithmic slope @xmath47 at @xmath48 1 mjy , steepening to @xmath49 at the bright end ( white 1996 , te96 ) just as one would expect if the brightest sources are at low redshifts @xmath50 where evolutionary and cosmological effects are negligible , and the same qualitative behavior is found at higher frequencies ( windhorst 1985 , 1993 ) .
using , we thus find that the expected number @xmath51 of sources above the flux cut is 4_^()d ( ) ^-(-1 ) , since the integral is dominated by sources just above the flux cut for which the power law fit is very accurate .
similarly , the integral in is dominated by sources just below the cut , so using equations and gives ( -13-)(n4)^2 ^3- .
finally , we can rewrite the lensing term in as ()^2 ^ 4 & & ( -1)^2(n4)^2 ^ 2 + & & ( -1)(3-)(n4 ) + & & ^-2(-2 ) . showed that that power spectrum was a sum of three different terms .
these are all plotted in as a function of the flux cut , using the source counts @xmath52 from the vla first survey ( white 1996 ) .
the approximations above allow us to understand all qualitative features of this figure .
comparing equations and , we notice that although the poisson term @xmath53 _ decreases _ if we remove more sources , the lensing term _ increases _ as long as @xmath46 , , as long as we remove less than @xmath54 sources from an all - sky survey , or @xmath55 sources per square degree .
this corresponds to @xmath56 sources in a @xmath57 field of the upcoming very small array ( _ vsa _ ) experiment , and to a 50 mjy flux cut at 1.5 ghz . shows that when @xmath58 , the ratio between the lensing term and the conventional shot noise term @xmath53 is simply @xmath41 times @xmath59 , the number of removed sources per steradian .
@xmath59 is typically @xmath60 , and we will now evaluate @xmath41 . for any flat friedman - robertson - walker cosmology , including non - linear clustering and an arbitrary source distribution @xmath61 , the magnification power spectrum is given by ( villumsen 1996 ) = 72 ^ 3_0 ^ 2 d^-3_0^x_h w^2(x_l ) ( 1+z)^2 p( / yd , z ) dx_l , where the function @xmath62 is defined as w(x_l)_x_l^x_h y_ls y_s n(x_s ) dx_s . here
@xmath63 , @xmath64 is the conventional three - dimensional matter power spectrum at redshift @xmath65 , and @xmath66 and @xmath67 are the comoving radial and angular distances to the epoch @xmath65 , respectively .
subscripts @xmath68 , @xmath69 and @xmath70 refer to horizon , lens and source positions , respectively , and @xmath71 is the comoving angular lens - source distance .
these equations are valid on angular scales where the sky is approximately flat , , @xmath72 , which is of course the case in our regime of interest . for the simple case where
@xmath73 , @xmath74 , the density fluctuations grow according to linear theory and all sources are located at some characteristic redshift @xmath75 , this reduces to = 72 ^ 3 d^-3 _ 0^x_s ( 1-xx_s)^2 p(xd)dx , where @xmath76 is the current power spectrum and @xmath77 .
.parameters used in optimistic , middle - of - the - road and pessimistic scenarios . [ cols="<,^,^,^",options="header " , ] this is plotted in ( dashed curves ) for a standard cdm power spectrum ( bond & efstathiou 1984 ) with @xmath78 , for the three cases described in table 1 .
we label these scenarios pessimistic " , middle - of - the - road " and optimistic " , since they are intended to give an upper limit , realistic estimate and lower limit , respectively , as to how much of a problem the lensing effect will turn out to be for future cmb experiments .
apart from a factor of @xmath79 , the magnification fluctuation on a given scale is essentially the area under a plotted curve out to that scale .
for instance , the magnification fluctuations after smoothing on a scale of 5 arcminutes is about @xmath80 for the pessimistic model , where the sources that have the dominant effect on the cmb are assumed to be at redshifts @xmath60 .
however , what matters for our purposes is of course not this magnification fluctuation , but @xmath41 itself . in the pessimistic scenario ,
the maximum power is @xmath81 , attained for @xmath82 , which ensures that the lensing fluctuations are below the standard poisson fluctuations for any reasonable flux cut .
this conclusion in independent of the overall normalization of the point source counts @xmath43 , since shows that the _ ratio _ of our lensing effect to the shot noise power @xmath53 is independent of this normalization . however , to assess the overall importance of both effects to future cmb experiments , accurate knowledge of the radio source counts at the frequencies where the observations are made is crucial .
unfortunately , the point source population between 20 and a few hundred ghz is still very poorly determined ( see gundersen 1997 ) , leaving open only the unattractive option of extrapolating from lower frequencies . in
we have extrapolated the 1.5 ghz vla first data to 30 ghz by assuming that the temperature fluctuations @xmath83 . to reflect the observational uncertainty , we have allowed @xmath84 to vary as shown in table 1 , with the middle - of - the - road estimate being that of te96 .
we have calculated the effect of weak gravitational lensing on the power spectrum of unresolved point sources .
we found that this adds a new term , which is the magnification power spectrum times @xmath39 .
the ratio between this term and the standard poisson term is roughly @xmath85 , independent of the ( poorly known ) point source normalization , so if we keep shrinking the point source region in by lowering the flux cut _ ad infinitum _ , the lensing term will eventually become the dominant problem .
the lensing term initially _
increases _ as we remove more sources , reaching a maximum when @xmath55 sources per square degree are removed .
shows that for high - resolution interferometric experiments operating at low frequencies , a fairly aggressive source removal scheme will be necessary , bringing the lensing effect close to this maximum strength .
effective point source removal will be important for the future satellite missions _
map _ and _ planck _ as well .
fortunately , this strength does not exceed 12% of the standard poisson fluctuations even in the worst - case scenario . in summary ,
although this effect should be taken into account in the detailed analysis of some future cmb experiments , it is likely to be a relatively small correction ( like , e.g. , the rees - sciama effect ) and should not substantially reduce the accuracy with which cosmological parameters can be measured . the authors wish to thank avi loeb for suggesting this project and for helpful comments on the manuscript .
this work was partially supported by european union contract chrx - ct93 - 0120 , by deutsche forschungsgemeinschaft grant sfb-375 and by nasa through a hubble fellowship , # hf-01084.01 - 96a , awarded by the space telescope science institute , which is operated by aura , inc . under nasa contract nas5 - 26555 .
bersanelli , m. 1996 , _ cobras / samba report on the phase a study _ ,
esa report d / sci(96)3 . the shaded regions indicate where the various foregrounds cause fluctuations exceeding those of cobe - normalized scale - invariant fluctuations , thus posing a substantial challenge to estimation of genuine cmb fluctuations .
they correspond to dust ( top ) , free - free emission ( lower left ) , synchrotron radiation ( lower left , vertically shaded ) and point sources ( lower and upper right ) .
the heavy dashed line shows the frequency where the total foreground contribution to each multipole is minimal .
the boxes roughly indicate the range of multipoles @xmath42 and frequencies @xmath86 probed by various cmb experiments , as in te96 .
[ everythingfig ] the three terms that contribute to the point source power spectrum in are plotted as a function of the number of sources removed in an all - sky survey ( only the number per steradian matters ) , based on the vla first source counts .
these terms ( solid curves ) are the integrated brightness of the population , the edge effect at the flux cut and the shot noise fluctuations , respectively .
the dashed curves show the lensing contribution to @xmath87 for the pessimistic and middle - of - the - road scenarios , at the worst - case @xmath42-values . | weak gravitational lensing from large - scale structure enhances and reduces the fluxes from extragalactic point sources with an amplitude of order 15% . in cosmic microwave background ( cmb ) experiments , sources exceeding some flux threshold @xmath0 are removed , which means that lensing will modulate the brightness map of the remaining unresolved sources . since this mean brightness is of order @xmath1 at 30 ghz for a reasonable flux cut
, one might be concerned that this modulation could cause substantial problems for future cmb experiments .
we present a detailed calculation of this effect and , fortunately , find that its power spectrum is always smaller than the normal point source power spectrum .
thus although this effect should be taken into account when analysing future high - precision cmb measurements , it will not substantially reduce the accuracy with which cosmological parameters can be measured . #
1#1 # 1;#2;#3;#4 # 1 , _ # 2 _ , * # 3 * , # 4 .
# 1;#2;#3;#4;#5 # 1 , _ # 2 _ , * # 3 * , # 4 ( # 5 " ) .
# 1@xmath2 # 1@xmath3 # 1equation ( [ # 1 ] ) # 1equation ( [ # 1 ] ) # 1 ( [ # 1 ] ) # 11 2.3truecm 11.45truecm#1 # 1figure [ # 1 ] # 1figure [ # 1 ] # 1to 0pt#1 # 1_#1 _ #
1#1 # 1#1 # 1#1 # 110^#1 march 4 , 1997 0.9truecm * * is lensing of point sources a problem for future cmb experiments?max / lensing.html _ ( faster from the us ) and from + _ h t t p://www.mpa - garching.mpg.de/@xmath4max / lensing.html _ ( faster from europe ) .
+ note that figures 1 and 3 will print in color if your printer supports it . ]
0.5truecm max tegmark@xmath5 & jens v. villumsen@xmath6 + + + 0.2truecm |
light rays are deflected as they travel through the universe , and as a result our view of the distant , faint galaxies is somewhat distorted . measuring these lensing induced distortions has proven to be a valuable tool in observational cosmology ( see the review by mellier 1999 ) : the lensing signal is sensitive to the projected mass surface density of the deflector , and through lensing one can probe the mass distribution without having to rely on assumptions about the dynamical state and nature of the deflecting matter .
when the surface density of the lens is sufficiently high , the distortions can be so large that multiple images can be seen , which provide strong constraints on the mass distribution in cluster cores . in this paper
we will concentrate on the weak lensing regime , which allows us to study the cluster mass distribution on larger scales .
weak lensing is a powerful tool to probe the distribution of matter in the universe on various scales .
one exciting application is the measurement of the lensing signal induced by large scale structure ( e.g. blandford et al .
1991 ; kaiser 1992 ; bernardeau , van waerbeke , & mellier 1997 ; schneider et al .
1998 , swjk hereafter ) , and recently several groups have reported detections of this signal ( bacon et al .
2000 ; kaiser , wilson , & luppino 2000 ; van waerbeke et al .
2000 ; wittman et al .
2000 ) .
the first successful weak lensing results were obtained by studying the systematic distortion of the images of faint galaxies behind rich clusters of galaxies ( e.g. tyson , wenk , & valdes 1990 ) .
measuring the masses and mass - to - light ratios of rich clusters is an important application of weak lensing , because comparison of the weak lensing result with other mass estimators ( e.g. , x - ray , s - z effect , or velocities ) can give important clues about the geometry and dynamical state of the system ( e.g , zaroubi et al .
1998 ) .
weak lensing also has some drawbacks .
first of all , the amplitude of the lensing signal depends on the redshift distribution of the faint background galaxies .
these galaxies are generally too faint to be included in spectroscopic redshift surveys , but photometric redshifts are thought to provide ample information on the redshift distribution of these galaxies ( e.g. , hoekstra , franx , & kuijken 2000 ) .
also the observed weak lensing signal is a measure of a surface density contrast , and one of the problems associated with this is the mass sheet degeneracy ( gorenstein , shapiro , & falco 1988 ) .
one could try to get around this problem by measuring the lensing signal out to large distances from the centre , and arguing that the cluster surface density should vanish there .
another potential problem is the fact that the lensing signal is sensitive to all matter along the line of sight . in general ,
weak lensing analyses of clusters assume that the signal induced by the cluster under investigation is so dominant that the contribution from other structures along the line of sight can be neglected .
the effect of structures , close to the cluster , such as the filaments that connect the cluster to the ` cosmic web ' , or an elongation along the line of sight , have been studied through n - body simulations ( e.g , cen 1997 , metzler et al .
local structures introduce both a bias and an additional uncertainty in the cluster mass estimate , in particular when one for instance tries to measure the mass within @xmath6 . in this paper
we study how distant structures affect weak lensing mass estimates of clusters of galaxies .
unlike the local structure , these are uncorrelated with the cluster , and they can be filaments , galaxy groups , or even other clusters . that such structures affect the mass estimates is easily seen in the case of a massive high redshift cluster , with a low redshift group in front of it . assume the cluster is at a redshift of unity , with a group at say @xmath7 , and that the mass of the group is one tenth of that of the cluster : the contribution of the group to the total lensing signal is about one third , compared to the 10% in mass . in this paper
we will quantify the uncertainty in weak lensing mass estimates of rich clusters caused by intervening matter .
to do so we examine the predicted statistical properties of the contribution by large scale structure .
we use an approach similar to what is done to predict the amplitude of the cosmic shear signal , which is used to constrain the cosmological parameters . given a cosmological model , we estimate the uncertainty in cluster masses derived from weak lensing analyses .
the structure of this paper is as follows . in section 2
we discuss two popular methods to measure cluster masses . in section 3
we derive the equations that describe the effect of large scale structure on such mass estimates .
the effect of large scale structure on mass estimates for clusters is quantified in section 4 .
we examine the dependence on cluster redshift , aperture size and source redshift distribution .
cluster masses can be measured in several ways . in this paper
we will use aperture mass statistics , which provide a filtered measurement of the of the surface mass density in an aperture ( schneider 1996 , s96 hereafter ) : @xmath8 where @xmath9 is the weight or filter function .
provided that @xmath9 is compensated , i.e. @xmath10 the aperture statistic can be related to the tangential shear @xmath11 .
unfortunately @xmath11 can not be observed directly , but one can measure the distortion @xmath12 . in the weak lensing regime , where @xmath13 , one can use @xmath14 , which is what we do throughout this paper .
then we have @xmath15 here , @xmath16 is related to @xmath9 through : @xmath17 s96 has shown that the signal - to - noise ratio of the aperture mass measurement is optimal when @xmath18 is the same as the ( expected ) tangential shear as a function of @xmath19 .
we want to quantify the uncertainty in mass measurements of clusters , and therefore require that the aperture statistic has a clear physical meaning .
studies of cosmic shear ( see swjk ) use different choices for @xmath18 .
below we discuss two statistics that have been used to estimate cluster masses .
a popular model to describe the cluster mass distribution is the singular isothermal sphere ( sis ) @xmath20 where @xmath21 is the einstein radius .
the einstein radius ( in radians ) is related to the velocity dispersion through @xmath22 the tangential shear as a function of radius is simply given by @xmath23 the best fit sis model is obtained by minimizing @xmath24 with respect to @xmath21 , where @xmath25 are the individual measurements of the lensing signal at radius @xmath26 . the uncertainty in the measurement of the tangential shear ,
@xmath27 is given by @xmath28 with @xmath29 the uncertainty in the measurement of an individual galaxy , @xmath30 the number density of sources , and @xmath31 the width of the @xmath32th radial bin .
if we take the widths of all the bins the same , and assume that the number density of sources is uniform , minimizing with respect to @xmath21 gives @xmath33 for a continuous profile the summation can be replaced by an integral . as mentioned before one measures the distortion and not the shear , and equation [ eqre ] only works well when @xmath34 is small . also substructure in
the cluster core changes the average tangential distortion ( e.g. , hoekstra et al . 2000 ) . to come around these problems
one has to choose an inner radius for the fit , such that equation [ eqre ] is a good approximation .
if we assume that the tangential shear has been measured from a distance @xmath35 out to @xmath36 , the einstein radius satisfies @xmath37 comparison with equation [ eqmapq ] shows that the best fit sis model is an aperture mass measurement where @xmath18 is given by @xmath38 the corresponding weight function for the surface density @xmath9 is given by @xmath39 the mass profiles of many clusters that have been studied to date are consistent with a sis model .
thus the weight function suggested above is close to the optimal choice .
however , this choice for @xmath40 is not a continuous function .
a functional form which is continuous , yet similar is suggested by s96 , who discusses the use of aperture mass statistics to find mass concentrations in deep images .
another statistic which has been widely used is the @xmath41-statistic ( fahlman et al .
1994 ) : @xmath42 it gives the difference of the average dimensionless surface mass density in a circular aperture of radius @xmath43 and that in an annulus with @xmath44 .
the corresponding weight function @xmath40 is given by : @xmath45 this gives @xmath46 thus the @xmath41-statistic can be related to the tangential shear through : @xmath47 the sis model fit has only a simple physical meaning when the cluster mass profile is indeed isothermal .
the interpretation of the @xmath41-statistic is independent of the mass distribution , although the statistic is not optimal in terms of signal - to - noise ratio . under the assumption that one can estimate or minimize @xmath48 , the average surface density in the control annulus , one obtains a measurement of @xmath49 , and thus of the cluster mass .
one way to do so is to assume a model for the cluster surface mass density which is used to estimate the average surface density in the annulus ( e.g. hoekstra et al .
1998 ) . this way
the cluster mass estimate is less dependent on the assumed model compared to the sis model fit .
under the assumption that the cluster is the only lensing structure in the field , the aperture mass @xmath50 provides a direct measure of the cluster mass .
however , the weak lensing signal is sensitive to all matter along the line of sight , and other structures , which we will refer to as large scale structure , will introduce an additional signal : @xmath51 the aperture mass statistic measures a density contrast , and the expectation value for the contribution by uncorrelated large scale structure @xmath52 ( see also swjk ) .
thus on average , distant structures do not introduce a bias in cluster mass estimates , but an uncertainty in the cluster mass of . as a result
the effect of distant large scale structure can be treated as an additional ( statistical ) source of error .
its relative importance can be decreased by averaging the results for several clusters , i.e. measuring their average mass .
we note that on certain scales @xmath53 might be distributed skewly .
the effect of ` local ' ( or correlated ) large scale structure has been studied using n - body simulations ( e.g. , cen 1997 ; metzler et al . 1999 ) . in this case
either projection effects , or filaments connected to the cluster bias the weak lensing mass estimates towards somewhat higher values .
this becomes particularly important when one tries to estimate the enclosed mass within several mpc s from the cluster centre ( cen 1997 ) .
swjk have shown how one can relate the aperture mass statistic to the power spectrum of density fluctuations .
measuring for a number of random fields allows one to constrain the projected power spectrum . in this paper
the problem is reversed : given the power spectrum , what is the resulting ?
a detailed discussion about the use of aperture mass statistics and cosmic shear can be found in swjk and bartelmann & schneider ( 1999 ) .
swjk showed that the effect of large scale structure on the variance in the aperture mass statistic is given by : @xmath54 where the effective projected power spectrum @xmath55 is given by : @xmath56 here @xmath57 is the radial coordinate , @xmath58 the cosmic scale factor , and @xmath59 the comoving angular diameter distance .
@xmath60 is the source averaged ratio of angular diameter distances @xmath61 for a redshift distribution of sources @xmath62 : @xmath63 the function @xmath64 in equation [ eqmap ] depends on the filter function @xmath40 , and is given by : @xmath65 as has been demonstrated in jain & seljak ( 1997 ) and swjk it is crucial to use the non - linear power spectrum in equation 20 .
this power spectrum can be derived from the linearly evolved cosmological power spectrum following the prescriptions of hamilton et al .
( 1991 ) , and peacock & dodds ( 1996 ) .
the form of the function @xmath66 , which probes the projected power spectrum , depends on the choice of @xmath40 . for the sis model fit we take @xmath67 , which yields @xmath68.\ ] ] @xmath69 we will use a typical value for @xmath70 ( e.g. , hoekstra et al .
larger values are used when much substructure is found in the cluster core ( e.g. , hoekstra et al .
the resulting @xmath71 is presented in figure [ filter]a . for the @xmath41-statistic
we find @xmath72 where we have taken the inner radius @xmath73 .
a typical value for @xmath74 used in weak lensing mass estimates of clusters is @xmath75 , i.e. , the control annulus runs from the boundary of the region in which we want to determine the mass out to 1.5 times this radius . throughout this paper
we will use this value for @xmath74 . in figure
[ filter]b the filter function @xmath76 is shown ( taking @xmath77 ) . the projected power spectrum @xmath78 depends on the redshift distribution of the sources , and consequently we need to know , or assume , a redshift distribution for the sources .
we will use the photometric redshift distributions inferred from the northern and southern hubble deep fields ( fernndez - soto , lanzetta , yahil 1999 ; chen et al .
1998 ) , which generally work well ( e.g. , hoekstra , franx , & kuijken 2000 ) .
there is , however , some field to field variation in the redshift distribution the sources , but we assume that the distributions obtained from the hdfs are representative for the whole sky .
we note that most of the lensing by large scale structure is caused by structures at intermediate redshifts , and as a result the signal induced by the lss is fairly robust against variations in the source redshift distribution . using the colours of the galaxies in the hdfs
we have computed their @xmath36 magnitudes , which allows a direct comparison with ground based observations .
we use galaxies fainter than @xmath79 .
the resulting photometric redshift distributions for different limiting magnitudes are presented in figure [ zdist ] .
.parameters of the cosmological models considered here .
all three models are cluster normalized.[cosmotab ] [ cols="<,^,^,^,^ " , ] given the photometric redshift distribution discussed above we calculate the contribution of the lss to the measurements of @xmath21 and the @xmath41-statistic as a function of aperture size . we have considered three different cosmologies , which are listed in table [ cosmotab ] .
figure [ mapscale]a shows the scatter in the value of the einstein radius @xmath21 as a function of aperture size for the scdm ( thick lines ) and the ocdm ( thin lines ) models . for clarity ,
we limit the number of curves shown in figure [ mapscale ] and omit the results for the @xmath80cdm model . for this model
the scatter is similar to the ocdm model , with a somewhat higher amplitude on large scales and a somewhat lower amplitude on arcminute scales .
the scatter in the value of @xmath21 increases for larger aperture sizes .
the value of @xmath21 is independent of aperture size , and as a result the relative contribution of large scale structure to the error budget increases with increasing aperture size . also the scatter is larger for fainter limiting magnitudes , i.e. higher source redshifts . figure [ mapscale]b shows the scatter for the @xmath41-statistic . because both @xmath81 and @xmath82 decrease with increasing aperture size the interpretation of figure [ mapscale]b is not as straightforward as is the case for figure [ mapscale]a .
the effect of aperture size on the accuracy of weak lensing mass estimates is discussed in detail in section 4.3 .
comparison of the results for the scdm and ocdm cosmologies indicates that the differences are fairly small .
we found that the @xmath80cdm cosmology gives similar estimates .
although we consider these different cosmologies throughout the paper ( and plot the results for the ocdm and scdm models ) we note that the results hardly depend on the assumed cosmology .
we examine the uncertainty in the weak lensing mass estimate caused by uncorrelated large scale structure .
the results are derived for a cluster with a sis mass profile , and a velocity dispersion of @xmath1 km / s , typical for rich clusters that have been studied in the past .
we study the dependence on cluster redshift , where we discuss the results for low redshift clusters in more detail .
we also examine how the accuracy of the mass measurements depends on aperture size .
first we examine how the mass estimates for clusters at different redshifts are affected when the lensing signal is measured out to a radius of @xmath2 mpc . for the sis model
we use the measurements in an annulus from 0.225 to @xmath2 mpc ( @xmath83 mpc , @xmath70 ) . in the case of the @xmath84statistic
we use an inner radius of @xmath85 mpc and an outer radius of @xmath86 mpc @xmath87 .
the value of the @xmath41-statistic only provides a lower limit to the cluster mass , and one way to circumvent this is to estimate the average surface density in the annulus around the aperture , for instance by using the results from a model fit .
however , this model fit is also affected by large scale structure . in general the region where the model is fitted to the data
largely coincides with the annulus used to compute the value of the @xmath41-statistic . thus if the value of the @xmath41-statistic is increased by additional structures
, the mass from the model will also be higher .
although the uncertainty in the model parameters and the @xmath41-statistic are not completely correlated , we will assume they are .
this gives a fair upper limit on the uncertainty in the estimate of @xmath88 @xmath89 the results are presented in figure [ ratio ] .
panels c and f show respectively the value of the einstein radius and @xmath88 as a function of cluster redshift .
we show the results for the scdm ( thick lines ) and the ocdm ( thin lines ) cosmologies .
the solid lines are for a limiting magnitude of @xmath90 and the dashed lines for @xmath91 . panels b and e
show the scatter in the aperture mass statistics introduced by the large scale structure .
finally the ratios of the scatter and the lensing signal are presented in figure a and d. for clusters at low redshifts @xmath92 the relative importance of large scale structure for the @xmath80cdm model is similar to that of the scdm model . for clusters at high redshift , or deep observations , the results for the @xmath80cdm are close to the ocdm results .
comparison of figures [ ratio]a and d shows that the relative importance of the distant large scale structure is smaller in the case of the sis model fit . for this mass estimator
the uncertainty in the cluster mass estimate is about 6% for rich clusters at intermediate redshifts .
the uncertainty in the mass is large for high redshift clusters when the limiting magnitude of the sources is rather bright ( solid lines ) .
however , in practice , to detect a significant lensing signal of a high redshift cluster deep observations are needed to ensure a sufficiently high number density of sources . as a result
the importance of large scale structure is minimized ( dashed lines ) . in the case of nearby clusters ,
the presence of structure along the line of sight results in a relatively large uncertainty in the cluster mass ( see section 4.2 ) .
the uncertainty introduced by large scale structure is much larger when the @xmath41-statistic is used .
this is not surprising as one only uses the shear measurements at relatively large radii , where the cluster mass surface density has become fairly low .
in addition to this annulus , the sis model fit uses data at much smaller radii , where the cluster signal is much higher . the noise in weak lensing analyses is generally dominated by the intrinsic shapes of the source galaxies , and not by the noise in the large scale structure .
both sources of error depend differently on the limiting magnitude of the background galaxies , and as a result the total noise is a rather complex function of cluster redshift and limiting magnitude , as is demonstrated in figure [ contsn]a for the ocdm cosmology . figure [ contsn]b shows the the signal - to - noise ratio for a sis model fit to the tangential shear of a cluster of @xmath1 km / s , when the contribution of large scale structure is ignored .
it should be noted here that we assume that the shapes of the sources have been measured accurately . in reality
the shapes of very faint galaxies are difficult to determine . from figure [ contsn]b
one would expect to be able to measure accurate masses for low redshift clusters .
however , the results change significantly when we include the contribution from large scale structure as is shown in figure [ contsn]a .
the change in signal - to - noise ratio is small for high redshift clusters , where the shapes are still the major contributor to the noise .
the true signal - to - noise ratio is lowered significantly in the case of deep observation of clusters at intermediate redshifts , but the largest difference is seen for clusters at low redshifts .
the signal - to - noise ratios for the @xmath80cdm model are intermediate to those found for the ocdm and scdm models , resembling the ocdm better for high redshift clusters .
however , at high redshifts the differences between the various models become less important ( they are fairly small anyway ) , because the noise is dominated by the intrinsic shapes of the sources .
figure [ contsn]a demonstrates that large scale structure introduces a large uncertainty in the mass measurement of low redshift clusters . to date
not many low redshift clusters have been studied through weak lensing .
this is because the amplitude of the lensing signal peaks at intermediate redshifts , and the signal is extremely small for nearby clusters .
furthermore their angular sizes were too large for the ccd cameras .
panoramic ccd cameras have become available recently , and a number of low redshift clusters have been imaged for weak lensing purposes .
for instance joffre et al .
( 2000 ) have analysed the mass distribution of abell 3667 @xmath93 .
although the lensing signal of low redshift clusters is small , there are several reasons to study these systems .
first of all , they can , or have been studied extensively by other techniques . also , the uncertainty in the redshift distribution of the sources is almost non - existent , and one can use relatively large background galaxies . in this section
we examine how the weak lensing mass measurement of the coma cluster @xmath94 is affected by structures at higher redshifts
. a weak lensing analysis of the most nearby rich cluster of galaxies is challenging , but could provide important insights in the cluster dynamics .
we assume a velocity dispersion of 1000 km / s for the cluster ( e.g. , colles & dunn 1996 ) . in figure [ noise_sis]a we plot the sources of noise in the measurement of the einstein radius as a function of limiting magnitude of the sources .
the figure shows the contribution from the large scale structure ( dashed lines ) and the intrinsic shapes of the sources ( dotted lines ) .
for the intrinsic shapes of the sources we have assumed @xmath95 , similar to what is observed in deep hst images ( e.g. , hoekstra et al .
2000 ) reaching the depth of the hdfs one would expect to obtain a signal - to - noise ratio of more than 20 .
figure [ noise_sis]a demonstrates that the total uncertainty in the measurement of the einstein radius reaches a minimum value for a limiting @xmath90 . figure [ noise_sis]b shows the expected signal - to - noise ratio as a function of limiting magnitude , and it demonstrates that observations with @xmath96 no longer improve the signal - to - noise ratio of the mass measurement .
the maximum signal - to - noise ratio that can be reached by fitting a sis model to the observed tangential shear around the coma cluster is @xmath97 for the scdm model , and @xmath5 for the ocdm model .
the results for the @xmath80cdm model lie between the curves for the other two models . in figure [ noise_zeta ]
we show how the accuracy of the mass measurement is affected by large scale structure when the @xmath41-statistic is used .
comparison with figure [ noise_sis ] shows that the sis model fit yields a better signal - to - noise ratio .
the intrinsic shapes of the sources result in a statistical uncertainty in the measurement of the einstein radius .
if the sis model is fitted to the measurements in an annulus from @xmath35 out to @xmath36 , the error in @xmath21 is given by @xmath98 where @xmath30 is the number density of sources , and @xmath29 is the uncertainty in the measurement of the tangential distortion for a single galaxy . from equation [ noisere ] we find that this contribution to the error budget decreases slowly for increasing aperture sizes .
however , figure [ mapscale]a shows that the noise contribution from large scale structure increases rather rapidly with aperture size . figure [ noise_app_sis]a shows how the contributions of the intrinsic shapes of the sources , and the large scale structure depend on the aperture size .
the results are for a cluster at a redshift @xmath99 , and the lensing signal has been measured from @xmath100 mpc out to a radius @xmath101 .
the uncertainty in @xmath21 caused by the intrinsic shapes of the sources decreases slowly with aperture size ( dotted curve ) , whereas the contribution by the distant large scale structure increases ( dashed line ) .
figure [ noise_app_sis]b shows the resulting signal - to - noise ratio .
the signal - to - noise ratio is virtually constant for outer radii larger than @xmath102 mpc .
measuring the lensing signal in larger apertures does not improve the accuracy of the cluster mass .
one has also to keep in mind that we have assumed that the mass profile is well described by a sis model , which might not be true in the outer parts of clusters . for clusters at higher redshifts
the curves look similar , but the maximum signal - to - noise ratio is reached at somewhat larger ( physical ) aperture sizes .
the noise caused by the large scale structure dominates the error for low redshift clusters , and for these systems , the signal - to - noise ratio decreases slowly with increasing aperture size .
so far we have used an inner radius for the sis model fit of @xmath100 mpc , but one can not always fit the model at such small distances from the centre .
for instance in the case of the @xmath103 cluster ms 1054 - 03 hoekstra et al .
( 2000 ) showed that the mass distribution in the centre is rather complex and the substructure lowers the azimuthally averaged tangential shear .
consequently they fit the sis model from @xmath104 out to @xmath105 mpc , which corresponds to @xmath106 .
we also have examined how the accuracy of the mass estimate changes when the sis model is fitted to the data within a given aperture , while the inner radius is increased .
in this situation both the noise caused by the shapes of the sources and the contribution by the large scale structure increase . thus small values for the inner radius improve the signal - to - noise of the mass measurement , but one should keep in mind that substructure in the cluster centre complicates a safe choice for the inner radius .
for the @xmath41-statistic the noise contribution caused by the shapes of the galaxies is given by @xmath107 however , we want to estimate @xmath108 , and thus we have to add the average surface density in the annulus from 1 to @xmath101 @xmath109 mpc . as before we estimate @xmath110 from a sis model fit to the tangential shear .
the errors in the @xmath84statistic and the einstein radius from the model fit are combined according to equation 25 , because the errors are correlated .
we note that the size of the correction depends on the assumed mass profile , but for increasing values of @xmath101 the correction becomes smaller ( i.e. , @xmath110 becomes smaller ) .
the resulting noise estimates are presented in figure [ noise_app_zeta]a .
the corresponding signal - to - noise ratio of @xmath108 for a cluster at a redshift @xmath99 is presented in figure [ noise_app_zeta]b .
it shows that using a larger @xmath101 improves the accuracy in the measurement of @xmath108 .
a similar calculation shows that the contribution of the large scale structure becomes more important when one tries to measure the average surface density in larger apertures .
for instance the uncertainty in @xmath111 almost doubles compared to @xmath108 .
attempts to measure the average cluster mass profile out to large projected distances from the cluster centre will also be affected by large scale structure , given the fact that it is already difficult to measure the total mass .
unfortunately it will be difficult to quantify the effect of distant large scale structure on such measurements .
the observed weak lensing signal is sensitive to all matter along the line of sight , and as a result this additional matter acts as a source of uncertainty in mass estimates of clusters of galaxies . in this paper
we have investigated the effect of distant large scale structure , uncorrelated with the cluster mass distribution .
it can be treated as an additional source of noise .
given a cosmological model it is possible to compute the statistical uncertainty in the weak lensing mass measurement caused by the distant large scale structure , and quantify its contribution to the total error budget .
we have considered two methods to estimate the mass from a weak lensing analysis .
one way is to fit a sis model to the observed lensing signal .
the other method uses the @xmath41-statistic ( e.g. , fahlman 1994 ) to estimate the mass .
both methods give similar results .
we find that the importance of matter along the line sight is fairly small for rich clusters ( @xmath1 km / s ) at intermediate redshifts , provided that the bulk of the sources are at high redshift compared to the cluster .
if the lensing signal is measured out to @xmath2 mpc from the cluster centre , the typical @xmath112 relative uncertainty in the mass cause by the large scale structure is about @xmath4 . in certain situations
the effect of large scale structure is more important . for nearby clusters , such as the coma cluster ,
background structures introduce a considerable uncertainty in the mass .
considering only the noise caused by the shapes of the sources used in the weak lensing analysis one expects to achieve very accurate weak lensing mass measurements of nearby clusters , but we find that the large scale structure limits the maximum achievable signal - to - noise ratio to @xmath5 , irrespective of the depth of the observations .
we also have examined how the accuracy of the weak lensing mass estimate changes when the lensing signal is measured out to large projected distances from the cluster centre . in the case of
the sis model fit we find that the signal - to - noise ratio does not increase significantly when one includes measurements at radii larger than @xmath102 mpc .
when the @xmath41-statistic is used to estimate the mass within a @xmath113 mpc radius aperture the signal - to - noise ratio does increase with @xmath101 .
if the cluster mass profile well described by a sis model , the model fit results in a better signal - to - noise ratio .
the effect of large scale structure is important in the outskirts of clusters , where the cluster signal itself is low . in this paper
we have quantified the uncertainty in the measurement of the cluster mass , but distant large scale structure will also introduce noise in studies that try to constrain the mass profiles of clusters at large projected distances .
furthermore studies of the 3-d structure of clusters by combining weak lensing , x - ray , and s - z data ( e.g. , zaroubi et al . 1998 ) will be affected by uncorrelated large scale structure .
photometric redshifts of the sources can be used to decrease the effect of large scale structure , either by modeling the contribution of large scale structure to the lensing signal , or by selecting sources in an optimal redshift range .
this can be only done at the expense of a large amount of additional data . because the effect of large scale structure can be treated as an additional ( statistical ) source of noise .
thus an alternative procedure is to average the results of an ensemble of clusters , thus decreasing the noise on the ( average ) cluster mass estimate .
we thank the referee for several useful comments which improved the paper .
it is a pleasure to thank marijn franx and konrad kuijken for useful discussions and comments to improve the paper .
also the comments from ludo van waerbeke are acknowledged .
peter schneider made available his algorithms , which were very helpful .
van waerbeke , l. , mellier , y. , erben , t. , cuillandre , j.c .
, bernardeau , f. , maoli , r. , bertin , e. , mc cracken , h.j . ,
le fevre , o. , fortm b. , dantel - fort , m. jain , b. , & schneider , p. 2000 , a&a , 358 , 30 | we quantify the uncertainty in weak lensing mass estimates of clusters of galaxies , caused by distant ( uncorrelated ) large scale structure along the line of sight .
we find that the effect is fairly small for deep observations @xmath0 of massive clusters ( @xmath1 km / s ) at intermediate redshifts , where the bulk of the sources are at high redshifts compared to the cluster redshift .
if the lensing signal is measured out to @xmath2 mpc the typical @xmath3 relative uncertainty in the mass is about @xmath4 .
however , in other situations the induced uncertainty can be larger .
for instance , in the case of nearby clusters , such as the coma cluster , background structures introduce a considerable uncertainty in the mass , limiting the maximum achievable signal - to - noise ratio to @xmath5 , even for deep observations .
the noise in the cluster mass estimate caused by the large scale structure increases with increasing aperture size , which will also complicate attempts to constrain cluster mass profiles at large distances from the cluster centre .
however , the distant large scale structure studied here can be considered an additional ( statistical ) source of error , and by averaging the results of several clusters the noise is decreased . |
recent studies on the molecular - gas properties in galaxies have revealed an intimate relationship between starbursts and high - density ( @xmath14 @xmath15 ) molecular gas ; quantitative correlations between hcn(1@xmath00 ) emission , which is an indicator of dense molecular gas contents , and star - formation tracers , such as a far - infrared ( fir ) continuum and radio recombination lines , have been demonstrated ( e.g. , solomon et al .
1992 ; zhao et al . 1996 ; paglione et al .
high - resolution images of hcn(1@xmath00 ) emission with millimeter - wave arrays also show spatial coincidence of dense molecular material with the massive star - forming regions in starburst / star - forming galaxies ( e.g. , paglione et al .
1995 ; kohno et al .
1999 , but see aalto et al .
2001 ) . given the important role of dense molecular matter on massive star- formation , it would be then important to investigate how they are formed .
a nonaxisymmetric distortion of the underlying potential in galaxies caused by various dynamical effects , such as disk instability and galaxy galaxy interactions ( e.g. , noguchi 1996 ) , is often invoked as an efficient mechanism which stuff interstellar matter ( ism ) on the large scale ( from a few to a few @xmath1 10 kpc ) disk of galaxies into the very central ( a few @xmath1 100 pc ) regions by removing the angular momentum of ism ( e.g. , shlosman et al .
1989 ) . to date , however , the physical process which could transform _ diffuse _ ism into dense molecular gas has not been addressed so often .
moreover , most of the observational studies on dense molecular gas have concentrated on luminous starburst galaxies , such as ngc 253 and m 82 .
it must be valuable to assess the distribution and physical properties of molecular gas in _
quiescent _ and _ less star - forming _ galaxies in order to seek a clue on dense gas formation . in this paper
, we report on high - resolution aperture synthesis co(1@xmath00 ) observations and simultaneous co(1@xmath00 ) as well as hcn(1@xmath00 ) spectroscopy of the early - type barred galaxy ngc 5195 , made with the nobeyama millimeter array ( nma ) and nro 45 m telescope .
as described below , ngc 5195 may be one of the ideal targets to investigate the molecular - gas properties in the galaxy whose star formation is currently at a very low level ; ngc 5195 is a barred lenticular galaxy ( sb0 pec ; sandage , tammann 1981 ) at a distance of 9.3 mpc ( tully 1988 ) , known as a companion of the well - studied spiral galaxy m 51/ngc 5194 . in spite of the abundant molecular gas in the central region of ngc 5195 ( sage , wrobel 1989 ; sage 1989 , 1990 ; aalto , rydbeck 2001 ) , the current massive star - formation seems to be suppressed there .
narrow - band imaging does not show any significant h@xmath16 _ emission _ , and strong h@xmath16 _ absorption _ dominates in the central @xmath17 region instead ( thronson et al .
1991 ; sauvage et al . 1996
; greenawalt et al . 1998
; see also optical spectra by filippenko , sargent 1985 ) .
this balmer absorption feature is attributed by a presence of numerous a - type stars , and few stars earlier than a4 exist in this region ( rieke 1988 ; yamada , tomita 1996 ) . considering the lifetime of a - type stars ( from a few @xmath1 @xmath18 to @xmath19 yr ) and the typical duration of starbursts ( a few @xmath1 @xmath20 ; thornley et al .
2000 and references therein ) , it is strongly suggested that ngc 5195 experienced a nuclear starburst about @xmath21 1 gyr before , and is now in the _ post - starburst phase _
, where the ob stars produced by the starburst event have disappeared and only late - type stars ( a stars and later ) remain in this region .
this situation is strikingly similar to the archetypical post - starburst galaxy ngc 4736 ( pritchet 1977 ; rieke et al . 1988 ; walker et al .
1988 ; taniguchi et al .
1996 ) . a mid - infrared ( mir ) study with isocam by boulade et al .
( 1996 ) has also revealed the post - starburst nature of ngc 5195 .
consequently , the current star - formation rate ( sfr ) is very small ; the h@xmath16 emission luminosity can be found by subtracting the underlying absorption feature by a stars , and is estimated to be @xmath22 erg s@xmath23 within a @xmath24 aperture ( ho et al .
this is near the lowest end of @xmath25(h@xmath16 ) in their sample , containing about 500 nearby galaxies in the northern hemisphere .
note that optical spectroscopy suggests the nucleus of ngc 5195 may be a liner ( e.g. , ho et al .
1997a ) , which could be a signature of a low - luminosity active galactic nucleus ( agn ) .
radio and mir observations also show a compact and luminous source in the center of ngc 5195 ( van der hulst et al .
1988 ; boulade et al .
it is , however , unclear whether a compact non - stellar powered source exists in the nucleus of ngc 5195 , because neither a compact uv core ( barth et al . 1998 ) nor ionized neon lines in the mir band ( boulade et al .
1996 ) , both are characteristic features of an agn , has been detected .
high - angular resolution x - ray observations of ngc 5195 also reveal an extended distribution of hot gas ( ehle et al . 1995 ;
georgantopoulos et al .
2002 ) , in favour of a star - forming origin for the bulk of the x - ray emission .
the radio and mir concentrations , therefore , would be related to past starburst events ( snrs and hot dust heated by late - type stars ) .
the parameters of ngc 5195 are listed in table 1 .
we performed simultaneous hcn(1@xmath00 ) and co(1@xmath00 ) observations toward the center of ngc 5195 using the nro 45 m telescope on 1996 december . in order to avoid the degradation of any beam pattern , beam efficiency , or pointing accuracy of the telescope ,
the observed data were discarded if the wind velocity exceeded about 4 m s@xmath26 .
the full widths of the half - power beam ( fwhp ) were @xmath27 and @xmath28 at the observed frequencies of co and hcn , respectively .
the main - beam efficiency , @xmath29 , was about 0.5 at the 3 mm band .
we used two cooled sis mixer receivers , s80 and s100 , equipped with side - band rejection filters , and observed both the co and hcn lines simultaneously .
the beam squint of the two receivers had been aligned to less than 2@xmath30 .
absolute pointing of the antenna was checked every hour using sio maser sources , and the pointing accuracy was better than @xmath31 ( peak - to - peak ) .
the typical system noise temperature for co and hcn observations were about 800 k and 300 k in the single side - band , respectively .
the sky emission was subtracted by position switching with an off - source position at an offset in the azimuth of @xmath32 from the observed position .
the spectra of co and hcn emission were obtained with 2048 channels acousto - optical spectrometers of 250 mhz bandwidth , corresponding to a velocity coverage of 650 km s@xmath26 for co observations , and 850 km s@xmath26 for hcn observations . after removing linear baselines from the spectra
, adjacent channels were averaged .
the resultant velocity resolution is 10 km s@xmath23 for co and 30 km s@xmath23 for hcn .
the 45 m telescope observations are summarized in table 2 .
nma observations of co(1@xmath00 ) emission in the central region of ngc 5195 were made during the period from 1999 november to 2000 march .
the nma consists of six 10 m antennas equipped with cryogenically cooled sis mixer receivers .
three available array configurations ( ab , c , and d ) were used . due to
the limitation of the minimum projected baseline length ( 10 m ) , extended structures larger than about @xmath33 in each channel map were not sampled in the observations .
the receiver noise temperatures at the 3 mm band were @xmath34 k in the double side - band , and the system noise temperatures in the single side - band were about 800 k toward the zenith .
the ultra wide - band correlator ( uwbc ; okumura et al .
2000 ) was configured to cover 512 mhz with 256 channels per baseline .
side - band separation was achieved by 90@xmath35 phase switching .
we observed the quasar 1418 + 546 every @xmath21 30 min in order to calibrate the temporal variations of the visibility amplitude and phase .
the passband across 256 channels was calibrated through observations of a strong continuum source , 3c 279 .
the flux density of the reference calibrator was determined from comparisons with planets of known brightness temperatures .
it ranged from 0.5 to 0.7 jy during the observation period , and the uncertainty in the absolute flux scale was about @xmath36 20 % .
the raw data were calibrated and edited using the package uvproc - ii developed at nro ( tsutsumi et al .
1997 ) , and fourier - transformed using the aips task imagr .
we produced two co data cubes with different imaging parameters ; one was maps using a natural weighting by applying a 80 k@xmath37 taper to the visibilities , and longer baseline visibilities ( @xmath38 k@xmath37 ) were cut .
this resulted in a moderate spatial resolution of @xmath1 ( p.a .
= @xmath0 @xmath39 ) .
the other cube was made using a robust weighting ( robust parameter = 0 ) to achieve a high angular resolution ( @xmath1 ) .
hereafter , we refer to these two cubes as `` low resolution '' and `` high resolution '' , respectively .
a conventional clean method was applied to deconvolve the synthesized beam pattern .
summaries of the observations and the parameters of the co data are listed in table 3 .
we detected strong co(1@xmath00 ) emission in the center of ngc 5195 , whereas hcn(1@xmath00 ) emission was found to be very weak .
the co and hcn spectra are shown in figure 1 , and the quantities derived from these spectra are summarized in table 2 .
the co profile shows two velocity features , i.e. , a @xmath40 km s@xmath23 component and a @xmath41 km s@xmath23 component , which are mostly consistent with the earlier co detection ( sage 1990 ) .
it has been suggested that the lower velocity component is from the spiral arm of m 51 , a companion of ngc 5195 , and that the higher velocity component is associated with ngc 5195 itself . a high - resolution nma map of co(1@xmath00 ) emission , presented in the following section , confirms this picture .
the velocity - integrated co intensity , @xmath42(co ) = 83 k km s@xmath23 , corresponds to a face - on molecular gas surface density of about 430 @xmath6 pc@xmath7 , adopting the formula @xmath43\ ] ] and @xmath44 where @xmath45 is the inclination of the disk ( @xmath46 here ) , @xmath47 is the velocity - integrated co intensity , and @xmath48 is the @xmath8 conversion factor ( scoville et al . 1987 ; solomon et al .
1987 ) .
we find that the hcn emission in the center of ngc 5195 is very weak .
because the detection was about the 3 @xmath49 level , it should be regarded as tentative . the hcn - to - co integrated intensity ratio in brightness temperature scale , @xmath10 ,
is given as about 0.02 .
note that the beam sizes of co and hcn observations are slightly different ( @xmath27 and @xmath28 , respectively ) , and one should note that such a difference could cause an error in the estimation of the intrinsic gas properties .
figure 2 shows velocity channel maps of the low - resolution co data cube in the central @xmath50 ( corresponding to 2.3 kpc at @xmath2 = 9.3 mpc ) area of ngc 5195 .
co emission is clearly detected ( @xmath51 ) over the velocity range from 487 to 737 km s@xmath23 , and presumably detected even in the 471 and 752 km s@xmath23 channels . a velocity - integrated co map ( the high resolution data ) over the @xmath52 km s@xmath23 range
is shown in figure 3c , along with an intensity - weighted mean - velocity map in figure 3d .
co maps made from the low - resolution data cube are shown in figure 4 .
these 0th and 1st moment maps were produced using the aips task momnt , with a clip level of 1.5 @xmath49 in each channel .
we can immediately see a strong concentration of co emission toward the nucleus of ngc 5195 .
the azimuthally averaged radial distribution of molecular gas shown in figure 5 , which was calculated using the aips task iring , suggests that the peak gas mass per unit area including he and heavier elements , @xmath3 , reaches about @xmath53 pc@xmath7 , adopting the equations ( 1 ) and ( 2 ) . the @xmath3 derived here
is much higher than that from the 45 m data .
this is solely due to the dilution effect ; the beam filling factor of molecular emission is very small .
the co peak corresponds to the radio peak position , and is mostly associated with the h@xmath16 _ absorption _ region , which can be seen in figure 3b . in this region , a hot dusty ring has also been discovered by a b-15 @xmath54 m color map ( block et al .
1997 ) , and which could also be a reminiscent of a past starburst event .
the relationship between molecular gas and the past starburst in the center of ngc 5195 is discussed in section 4 . the co flux , @xmath55 , within the central @xmath56 pc ( @xmath57 ) region in figure 4 is 280 jy km s@xmath23 , and the total flux detected in the f.o.v .
is about 380 jy km s@xmath23 .
these correspond to a molecular gas mass of @xmath58 and @xmath59 , respectively , following the equations @xmath60\ ] ] and @xmath61 the averaged gas surface density within the @xmath56 pc region is @xmath62 pc@xmath7 .
this @xmath3 is comparable to those in a normal spiral survey ( sakamoto et al .
1999b ) , which consists of 20 co - luminous spiral galaxies with morphology types from sa to sd .
note that the number of s0 galaxies where high - resolution co images are available is still limited to date ( e.g. , ngc 4701 , wrobel , kenney 1992 ; ngc 404 , cepa et al .
1998 ; ngc 3593 , sakamoto et al . 1999a ; ngc 7465 , kohno et al .
2001 ) , though a large number of co detections have been reported ( e.g. , sage , wrobel 1989 ; thronson et al . 1989 ; tacconi et al . 1991
; taniguchi et al . 1994 ) .
our results suggest that some s0 galaxies can contain a large amount of molecular gas comparable to disk galaxies .
the co velocity field in the low - resolution map ( figure 4 ) seems to be dominated by circular rotation . a simple
fitting of the observed velocity field with the aips task gal gives a systemic velocity of about @xmath63 km s@xmath23 and a p.a .
of the kinematic major axis of about @xmath64 ( measured counterclockwise from north ) , although it can depend on the area where the fitting is made .
is mostly consistent with the value in rc3 ( table 1 ) . on the other hand ,
the systemic velocity derived here differs significantly from the h velocity listed in rc3 , @xmath65 km s@xmath23 .
our co measurements must be more accurate , because high - resolution observations of h do not show any significant h emission in the center of ngc 5195 ( rots et al .
1990 ) , and the listed h velocity in rc3 seems to be contaminated by the h from m 51 . a position - to - velocity map ( @xmath66 map ) along the kinematic major axis ( p.a .
= + 106@xmath35 ) is displayed in figure 6 .
we find a steep rise of rotation velocity on the @xmath66 map ; it reaches about 160 km s@xmath23 at @xmath67 or 45 pc on the plane of the galaxy , adopting @xmath68 .
note that we see a velocity component near @xmath69 km s@xmath23 , which may be co emission associated with m 51 ( see 3.2.2 ) .
the observed gas morphology is reminiscent of those in barred spiral galaxies ( e.g. , kenney et al .
1992 ; sakamoto et al . 1995 ; reynaud , downes 1997 ; regan et al .
1999 ) ; two offset ridges emanate from the central gas concentration , and are connected to curved structures . a curved or spiral - like feature , which can often be a signature of the gas orbit resonance in barred galaxies , is particularly evident in the nw part .
in fact , near - infrared ( nir ) photometry clearly shows the presence of a large - scale stellar bar along the n
s direction ( smith et al .
1990 ; thronson et al . 1991 ; spillar et al . 1992 ; block et al .
1994 ) ; it is very natural to have an idea that they govern the gas distribution .
if we assume that these are bar - induced structures , indeed , it is possible to specify the direction of rotation on the plane of the galaxy . in figure 7
, we suggest a possible configuration . as can be seen in the figure ,
the northern part of the galaxy is suggested to be the near side , which is consistent with results obtained by assuming the outer spiral arms seen in nir are trailing ( block et al .
1994 ) . in order to fully understand the gas distribution and kinematics in the center of ngc 5195 , detailed analyses on gas kinematics , including a comparison with the hot dusty ring ( block et al .
1997 ) , will be needed .
it should be noted that low - level ( @xmath70 @xmath49 ) emission can be seen near to the edge of the nma f.o.v .
( mostly the s
e portion of the f.o.v . ) in the channel maps of the low - resolution co cube ( figure 2 ) for the velocity range @xmath71 km s@xmath23 . if we correct the primary beam attenuation near the edge of f.o.v .
, these low - velocity emissions could be about 200 mjy beam@xmath23 or 1 k at the observing beam ( @xmath72 ) .
the velocity range of these low - level co emissions is mostly the same as that of the low - velocity component detected at 45 m ( and nrao 12 m by sage 1990 ) .
we show the integrated intensity map for the 378 to 440 km s@xmath23 range in figure 8 .
a peak near the s
e edge of the f.o.v . can be clearly seen in the figure , and this could be co emission from m 51 .
this low - level emission is clearly depicted by recent multi - pointing imaging of the ngc 5195 region using the ovro millimeter array ( aalto , rydbeck 2001 ) .
nevertheless , the flux ratio between 45 m and nma data is very small for this velocity range , as can be seen in figure 9 , which compares the 45 m and nma spectra , although most of the co emission from ngc 5195 itself seems to be recovered with the nma observations .
we therefore suggest that the co emission from m 51 seen as the low - velocity component would be spread over a wide area of arcminutes scale , and that most of the low - velocity co emission is resolved out in our interferometric observations .
this seems to be consistent with the @xmath66 map by sage ( 1990 ) , which shows extended emission over a few arcminutes for the velocity component near @xmath73 km s@xmath23 .
the nma observations of co emission reveal that the post - starburst galaxy ngc 5195 does contain a significant amount of molecular gas , despite the fact that the current massive star formation is very inactive there ; the molecular gas mass per unit area reaches about @xmath74 pc@xmath7 , which is comparable to those in nearby starburst galaxies . on the other hand , the star formation rate ( sfr ) per unit area , @xmath75 , which was calculated from @xmath24 aperture h@xmath16 data ( ho et al .
1997a ) , is only about @xmath76 yr@xmath23 pc@xmath7 ( table 1 ) . this sfr is smaller than those in typical starburst galaxies by an order of magnitude ( e.g. , kennicutt 1998a ) .
the resultant gas consumption time scale , @xmath77 is about @xmath78 yr , which is indeed comparable to @xmath79 in quiescent or normal galaxies , and significantly longer than those in infrared - luminous starburst galaxies ( e.g. , kennicutt 1998a ) . in the following section we discuss the possible relationship between the physical properties of molecular gas and star formation in the center of ngc 5195 .
we observed a very low @xmath10 value of 0.02 in the center of ngc 5195 . this is smaller than those in starburst galaxies by a factor of @xmath80 .
for instance , @xmath10 in ngc 253 is reported to be in the range of @xmath81 ( helfer , blitz 1993 ; paglione et al .
1995 ; sorai et al .
2000 ) , and starburst galaxies such as ngc 6946 and ic 342 , which are less active compared with ngc 253 in terms of fir luminosities , show @xmath10 values of about 0.1 ( downes et al . 1992 ; helfer , blitz 1997 ) .
@xmath10 values in the circumnuclear starburst regions of ngc 1068 ( helfer , blitz 1995 ) and ngc 6951 ( kohno et al .
1999 ) are also about 0.1 or so . because the critical gas density for collisionally excitation of hcn(1@xmath00 ) emission ( @xmath14 @xmath15 ) is much higher than that of co(1@xmath00 ) ( @xmath82 a few @xmath1 @xmath83 @xmath15 ) ,
a comparison of the co and hcn intensities is a measure of the gas density if both of the co and hcn emission originate from the same volume . in observations of galaxies ,
the observing beams are often too large to resolve the individual cloud structures , and @xmath10 could indicate the fraction of dense molecular gas to the total ( including diffuse ) molecular gas within the observing beam ( kohno et al .
1999 ) . with this interpretation of @xmath10 in galaxies , our data on ngc 5195 and its comparison with starburst galaxies
would indicate that _ the mass fraction of dense molecular components to the total molecular gas _ , which is traced by the @xmath10 values , _ in the center of ngc 5195 is significantly small compared with that of starburst galaxies_. we suggest that this would be a reason why ngc 5195 is in a post - starburst phase despite the fact that it contains a large amount of molecular gas there ; most of the dense molecular cores , which are the very sites of massive star formation , have been consumed away by starburst events which began about 1 gyr before ( see section 1 ) .
also , a burst of massive star formation can no longer last , although a large amount of _ diffuse _ molecular gas remains there .
we should note that the @xmath10 values often depend on the aperture size of the observations , and attention must be paid to derive the intrinsic properties of molecular gas from the @xmath10 values .
( e.g. , helfer , blitz 1993 ; jackson et al .
for instance , @xmath10 in the center of m 51 is 0.033 ( helfer , blitz 1993 ) by the nrao 12 m telescope ( @xmath84 for co , and @xmath85 for hcn ) , whereas it becomes about 0.08 ( sorai et al .
2002 ) by the nro 45 m telescope ( @xmath27 and @xmath28 for co and hcn , respectively ) . in higher resolution ( @xmath86 ) observations , it reaches about 0.3 or more ( kohno et al .
these effects are mostly due to the compactness of the hcn emitting regions , and therefore high - resolution observations are essential in order to address the relationship between dense molecular gas and massive star - forming regions . in the case of our ngc
5195 observations , the observing beams ( @xmath27 for co and @xmath28 for hcn ) correspond to aperture areas of @xmath87 pc and @xmath88 pc for co and hcn , respectively .
small @xmath10 values of less than 0.05 have been reported in various galaxies ( e.g. , helfer , blitz 1993 ; aalto et al .
1995 ) , but are mostly observed with a much larger beam size ( @xmath89 ) than ours , corresponding to a few @xmath1 kpc aperture area in their samples .
we therefore suggest that the observed small @xmath10 value in the central a few @xmath1 100 pc of ngc 5195 is not the aperture size effect but reflects the intrinsic property of molecular gas , i.e. , a decrease of dense molecular gas there .
our interpretation of a low @xmath10 ratio seems to be supported by a low co(2@xmath01)/co(1@xmath00 ) ratio in ngc 5195 ; sage ( 1990 ) observed co(2@xmath01 ) and co(1@xmath00 ) lines toward the center of ngc 5195 using the nrao 12 m telescope , and found that the co(2@xmath01)/co(1@xmath00 ) ratio is as small as 0.5 .
they concluded that the molecular clouds in ngc 5195 are _ less dense _ than those in typical spiral galaxies .
given the close relationship between dense molecular gas and massive star formation / starburst , what kind of physical processes can play a major role in the formation of _ dense _ molecular clouds from a _ low - density _ molecular medium ? one of the key processes could be gravitational instabilities of molecular gas ; it is well known that sfr in the disk regions ( @xmath21 a few kpc scale ) of galaxies can be described by means of the schmidt law , @xmath90 , where @xmath75 is the sfr per unit area , and @xmath3 is the gas mass per unit area ( kennicutt 1998a ) . here
it should be noted that the schmidt law is valid only if the gas in the disk is gravitationally _ unstable _ , i.e. , the gas surface density , @xmath3 , exceeds the critical gas surface density , @xmath91 , given by @xmath92 where @xmath16 is a dimensionless constant close to unity is unity . yet
@xmath16 could be less than unity for a realistic two - fluid stability condition because the interaction between the stellar and gaseous disks could act to destabilize the gas disk ( jog , solomon 1984 ) .
martin and kennicutt ( 2001 ) has shown that the best - fit value of @xmath16 is 0.69 @xmath36 0.2 ; we use this in the following calculation . ]
, @xmath93 is the velocity dispersion in the radial direction , and @xmath94 is the epicyclic frequency , @xmath95 \right\}^{0.5},\ ] ] where @xmath96 is the distance from the center and @xmath97 is the rotation velocity at the radius of @xmath96 after correcting the inclination as 1/sin @xmath45 . if the gas surface density , @xmath3 , exceeds the threshold , @xmath91 , then a uniform gas disk is unstable to form rings or clumps which can ultimately collapse into dense molecular gas fragments and form stars , eventually .
this criterion , for local gravitational stability in thin isothermal disks , is often expressed as @xmath98 ( toomre s @xmath99 parameter ; toomre 1964 ) ; @xmath100 ( @xmath101 ) means the gas disk is stable ( unstable ) . surprisingly , this criterion has been successfully applied to much smaller scale disks , i.e. , inner a few @xmath1 100 pc regions of galaxies , to explain the star - formation properties of starbursts ( e.g. , ngc 3504 , kenney et al .
1993 ) and rather normal ( e.g. , ngc 4414 , sakamoto 1996 ) galaxies .
it is therefore natural to expect that the gravitational instabilities of molecular gas would be related to dense molecular gas formation .
in fact , kohno et al .
( 1999 ) reported that @xmath10 is enhanced in regions where the @xmath99 parameter is almost equal to or smaller than unity , suggesting a connection between dense gas formation and gravitational instabilities of molecular gas . in order to assess the stability of a molecular gas disk in the center of ngc 5195
, we calculated @xmath91 in ngc 5195 using the above equations .
the circular rotation velocities , obtained from the @xmath66 map in figure 6 , were employed to compute the epicyclic frequency , @xmath94 .
note that our interest is just at the center of the gas disk , and that the gas motion can be treated as a rigid rotation in this region .
in this case , the epicyclic frequency can be written just as @xmath102.\ ] ] from the @xmath66 map in figure 6 , we suppose that the largest circular rotation velocity occurs at about @xmath67 or 50 pc and the @xmath103 is about @xmath104 km
s@xmath23 , yielding a @xmath94 of 4.5 within @xmath105 pc .
note that @xmath94 must be a lower limit because the radius of @xmath106 where the rotation rises to its peak could be an upper limit , given our limited angular resolution ( @xmath107 ) .
concerning the velocity dispersion of molecular gas , we here assumed an intrinsic gas velocity dispersion , @xmath93 , of 30 km s@xmath23 , which is similar to those in ngc 3504 ( kenney et al .
this is because a precise extraction of intrinsic gas velocity dispersion from the observed co line width ( figure 9 ) is not easy near the center of galaxies due to contamination of the steep velocity gradient within the observing beam .
consequently , we find a critical density , @xmath91 , of @xmath108 pc@xmath7 , and that the @xmath98 is larger than unity ( 1.9 ) in the center of ngc 5195 .
hence , the molecular gas disk in the center of ngc 5195 would be gravitationally _ stable _ ; we suggest that this may be the reason why dense molecular gas is not formed from the large amount of diffuse molecular gas in ngc 5195 , since they are _ too stable to form dense molecular clouds via gravitational instabilities of a diffuse molecular medium_. although there are some sources of error , we regard this conclusion as begin rather robust .
though one of the major uncertainties of the @xmath99 parameter comes from @xmath48 , recent studies tend to show _ smaller _ values than the @xmath48 we adopted here by a factor of @xmath109 or more in the central regions of various galaxies ( e.g. , nakai , kuno 1995 ; @xcite ) , including the galactic center ( e.g. , oka et al .
1998 ; dahmen et al . 1998 ) . even in the solar neighborhood ,
the latest extensive co survey data of the milky way suggest a @xmath48 of @xmath110 cm@xmath7 ( k km s@xmath23)@xmath23 ( dame et al .
2001 ) , which is indeed smaller than the @xmath48 adopted in our calculation .
thus the uncertainty of @xmath48 does not weaken our conclusion .
although another source of error is the velocity dispersion , a large velocity dispersion of up to a few @xmath1 10 km s@xmath23 can be observed even in rather quiescent galaxies ; the galactic center is the case ( e.g. , spergel , blitz 1992 ) . given the finite spatial resolution , the epicycle frequency , @xmath94 , could be a lower limit , indicating that the critical density can be even larger .
it increases the @xmath99 value then .
all of these error sources thus indicate that the molecular gas in the center of ngc 5195 is indeed supercritical ( @xmath100 ) despite the large molecular gas mass there .
supercritical ( @xmath100 ) molecular gas in post - starburst galaxies has also been reported at the centers of ngc 4736 ( shioya et al .
1996 ) and ngc 7331 ( tosaki , shioya 1997 ) ; both of them are prototypical post - starbursts ( see taniguchi et al .
1996 and reference therein for ngc 4736 , and see ohyama , taniguchi 1996 for ngc 7331 ) .
gravitational instabilities of molecular gas may , therefore , be closely related to dense molecular gas formation and , in turn , massive star - formation in galaxies .
it should also be addressed why the molecular gas disk in ngc 5195 is supercritical . as mentioned above ,
the gas surface density , @xmath3 , is rather comparable to those in nuclear starburst regions of galaxies .
the major difference between starbursts and the post - starburst galaxy ngc 5195 may be the very high critical density for the gravitational instability ; the deduced @xmath91 of @xmath108 pc@xmath7 is significantly high compared with the central gas surface density in disk galaxies ( e.g. , sakamoto et al .
1999b ) .
this very high threshold may be caused by the large mass concentration in early - type galaxies ( kennicutt 1989 ) .
the large concentration of matter results in a steep rise of the rotation velocity , as observed in ngc 5195 .
this increases the threshold density , because @xmath94 becomes very high in this case .
in other words , molecular gas at the centers of early - type galaxies tends to be stabilized due to a small gas mass - to - dynamical mass ratio , @xmath112 , even if a significant molecular gas mass concentration , comparable to late - type disk galaxies , occurs ( e.g. , young , scoville 1991 ) . these properties observed in the s0 galaxy ngc 5195 may explain the difference in the star - formation properties along the hubble sequence .
as reviewed by kennicutt ( 1998b ) , it is now established that the detection frequency of nuclear star formation is lower in early - type galaxies than in late - type ones ( e.g. , the detection rate is about 8% in s0 , whereas it increases about 50% in sb and reaches 80% in sc - im ; ho et al .
1997b ) , whereas the strength of nuclear star formation shows the opposite trend ( i.e. , the average extinction - corrected h@xmath16 luminosities in s0sbc galaxies is nine - times higher than in sc galaxies ; ho et al .
the high critical density due to the high mass concentration in early - type galaxies may explain these apparently incoherent trends ; in early - type galaxies , the occurrence of nuclear star formation is low because nuclear star formation can not occur until the gas surface density exceeds the higher threshold density compared with late - type galaxies . in order to exceed the threshold
, it may take a longer time or many triggers to transport the larger amount of molecular gas into the centers of early - type galaxies .
once the gas surface density exceeds the threshold , however , the star - formation rate must be very high , because the gas surface density already reaches a very high value compared with late - type galaxies ( kennicutt 1989 ) .
our high - resolution co observations of the early type galaxy ngc 5195 do support this scenario .
a co survey of early - type galaxies using millimeter - wave interferometers and a comparison of the gas surface and the threshold densities with those in late - type spirals must be valuable to statistically test the picture described above regarding the difference in star formation along the hubble sequence .
we present high - resolution co(1@xmath00 ) and hcn(1@xmath00 ) observations of the post - starburst galaxy ngc 5195 made with the nobeyama millimeter array and the nro 45 m telescope .
our results and conclusions are summarized as follows : 1 .
we detected strong co emission in the center of ngc 5195 , whereas hcn emission was found to be very weak .
the hcn - to - co integrated intensity ratio in brightness temperature scale , @xmath10 , is about 0.02 .
this @xmath10 is very small compared with starburst galaxies by a factor of about @xmath12 .
2 . our nma map of co emission reveals a concentrated co distribution toward the nucleus , where little massive star formation is observed ; the molecular gas mass per unit area reaches about @xmath74 pc@xmath7 , which is comparable to those in nearby starburst galaxies , whereas the massive star formation rate in the center is smaller than those in typical starburst galaxies by orders of magnitude .
the resultant gas consumption time scale , @xmath79 , is about @xmath78 yr , showing very small star - formation efficiency in the center of ngc 5195 compared with that of starburst galaxies .
we suggest that the very small @xmath10 value could explain the reason why the starburst in ngc 5195 has ceased ; most of the dense molecular cores , which are the very sites of massive star - formation , have been consumed away by starburst events which began about 1 gyr before .
a burst of massive star formation can no longer last , although a large amount of _ diffuse _ molecular gas remains there .
4 . we have found a steep rise of rotation velocity toward the center of ngc 5195 .
it reaches about 160 km s@xmath23 at @xmath113 pc on the plane of the galaxy . as a consequence ,
the critical gas surface density for gravitational instability of the gas disk is very high ( @xmath114 @xmath6 pc@xmath7 ) , and we suggest that the gas disk in the center of ngc 5195 is gravitationally stable .
although ngc 5195 contains large quantities of _ low density _ molecular gas in the center , they must be too stable to form dense molecular clouds via gravitational instabilities of molecular gas .
the deduced very high threshold density seems to be due to a high mass concentration in the center of ngc 5195 , and may be characteristic of early - type galaxies . if this is the case , the known trends on the occurrence and luminosity of nuclear star formation along the hubble sequence can be explained naturally .
we would like to thank the referee for the helpful comments that have improved this paper .
we acknowledge the nro staff for operating the telescopes and continuous efforts to improve the performance of the instruments .
aalto , s. , httemeister , s. , & walter , f. 2001 , in asp conf .
249 , the central kiloparsec of starbursts and agn : the la palma connection , ed .
j. h. knapen , j. e. beckman , i. shlosman , and t. j. mahoney ( san francisco : asp ) , 626 aalto , s. , & rydbeck , g. 2001 , in asp conf .
249 , the central kiloparsec of starbursts and agn : the la palma connection , ed .
j. h. knapen , j. e. beckman , i. shlosman , and t. j. mahoney ( san francisco : asp ) , 719 kohno , k. , matsushita , s. , vila - vilar , b. , okumura , s. k. , shibatsuka , t. , okiura , m. , ishizuki , s. , & kawabe , r. 2001 , in asp conf . ser .
249 , the central kiloparsec of starbursts and agn : the la palma connection , ed .
j. h. knapen , j. e. beckman , i. shlosman , and t. j. mahoney ( san francisco : asp ) , 672 ( astro - ph/0206398 ) @ p12pcccc + parameter & value & reference & + [ 4pt ] + morphology & ia0 pec & ( a ) + & sb0 pec & ( b ) + size ( @xmath115 ) & @xmath116 & ( a ) + nuclear activity & l2 : & ( c ) + & post - starburst & ( d ) , ( e ) + position of nucleus & & ( f ) + @xmath16(j2000 ) & @xmath117 & + @xmath16(j2000 ) & @xmath118 & + position angle & @xmath119 ( counterclockwise from n ) & ( a ) + inclination angle & @xmath120 ( face - on = @xmath121 ) & ( a ) + distance & 9.3 mpc & ( g ) + liner scale & 45.1 pc arcsec@xmath23 & + h@xmath16 luminosity ( @xmath122 area ) & @xmath123 erg s@xmath23 & ( c ) + sfr per unit area@xmath124 @xmath75 & @xmath125 @xmath126 pc@xmath7 yr@xmath23 & + @ p20pcccc + parameter & co(1@xmath00 ) & hcn(1@xmath00 ) & + [ 4pt ] + beamsize ( fwhp ) & @xmath27 & @xmath28 + main beam efficiency @xmath29 & 0.5 & 0.5 + velocity resolution @xmath128 [ km s@xmath23 ] & 10 & 30 + peak temperature in @xmath130 [ mk ] & @xmath131 & @xmath132 + integrated intensity@xmath124 @xmath133 [ k km s@xmath23 ] & @xmath134 & @xmath135 + luminosity @xmath136 [ k km s@xmath23 pc@xmath137 & @xmath138 & @xmath139 + @ p20pccc + parameter & value & + [ 4pt ] + field of view ( fwhp ) & @xmath142 + & 2.7 kpc at @xmath2 = 9.3 mpc + bandwidth & 512 mhz + visibility calibrator & 1418 + 546 + flux density of calibrator & @xmath143 jy + array configuration & ab , c , and d + + weighting & natural , 80 k@xmath37 taper + synthesized beam & @xmath72 , @xmath144 + & 210 pc @xmath1 140 pc + equivalent @xmath145 for 1 jy beam@xmath23 & 6.1 k ( jy beam@xmath23)@xmath146 + velocity reslution @xmath128 & 15.6 km s@xmath23 + rms noise level in channel maps & 23 mjy beam@xmath23 + co flux within the f.o.v . &
km s@xmath23 + molecular gas mass within the f.o.v.@xmath124 & @xmath148 @xmath6 + peak face - on gas surface density@xmath124 & @xmath149 @xmath6 pc@xmath7 + co intensity at @xmath27 beam@xmath150 & 93.2 k km s@xmath23 + @xmath42(nma)/@xmath42(45m)@xmath150 & 1.1 + + weighting & robust ( robustness 0 ) + synthesized beam & @xmath1 , @xmath151 + & 86 pc @xmath1 81 pc + equivalent @xmath145 for 1 jy beam@xmath23 & 27 k ( jy beam@xmath23)@xmath23 + velocity reslution @xmath128 & 15.6 km s@xmath23 + rms noise level in channel maps & 22 mjy beam@xmath23 + @xmath150 the nma co cube was convolved with @xmath27 beam ( the same as the 45 m observations ) and co intensity at the center ( the observed potition with the 45 m ) , integrated over the velocity range of @xmath153 km s@xmath23 , was compared with the 45 m data ( table 2 ) . most of the single dish flux seems to be recovered in this velocity range . | the nobeyama millimeter array ( nma ) has been used to make aperture synthesis co(1@xmath00 ) observations of the post - starburst galaxy ngc 5195 .
co(1@xmath00 ) and hcn(1@xmath00 ) observations of ngc 5195 using the nobeyama 45 m telescope are also presented .
high - resolution ( @xmath1 or 86 pc @xmath1 81 pc resolution at @xmath2 = 9.3 mpc ) nma maps show a strong concentration of co emission toward the central a few @xmath1 100 pc region of ngc 5195 , despite the fact that the current massive star formation is suppressed there . the face - on gas surface density , @xmath3 , within the @xmath4 or 90 pc region reaches @xmath5 @xmath6 pc@xmath7 if a galactic @xmath8 conversion factor is applied
. the extent of the central co peak is about @xmath9 , or 230 pc , and is elongated along the e w direction with two - armed spiral - like structures , which are typical for barred disk galaxies .
the hcn - to - co integrated intensity ratio on the brightness temperature scale , @xmath10 , is about 0.02 within the central @xmath11 pc region .
this @xmath10 is smaller than those in starburst regions by a factor of @xmath12 .
these molecular - gas properties would explain why ngc 5195 is in a post - starburst phase ; most of the _ dense _ molecular cores ( i.e. , the very sites of massive star formation ) have been consumed away by a past starburst event , and therefore a burst of massive star formation can no longer last , although a large amount of _ low density _ gas still exists .
we find a steep rise of the rotation velocity toward the center of ngc 5195 . as a consequence ,
the critical gas surface density for a local gravitational instability of the gas disk becomes very high ( @xmath13 @xmath6 pc@xmath7 ) , suggesting that the molecular gas in the central region of ngc 5195 is gravitationally _
stable _ , in contrast to that of starburst galaxies .
we propose that dense molecular gas can not be formed from remaining diffuse molecular gas because the molecular gas in the center of ngc 5195 is _ too stable _ to form dense cores via gravitational instabilities of diffuse molecular gas .
the deduced very high threshold density seems to be due to a high mass concentration in ngc 5195 .
the known trends on the occurrence and luminosity of nuclear star formation in early - type galaxies can be understood naturally if the high threshold density is characteristic for early - type galaxies . |
in this paper we consider semiparametric models defined by conditional mean and conditional variance estimating equations .
models defined by estimating equations for the first and second order conditional moments are widely used in applications .
see , for instance , ziegler ( 2011 ) for a recent reference . here
we consider a model that extends the framework considered by cui , hrdle and zhu ( 2011 ) . to provide some insight on the type of models we study ,
consider the following semiparametric extension of the classical poisson regression model with unobserved heterogeneity : the observed variables are @xmath0 where @xmath1 denotes the count variable and @xmath2 is the vector of @xmath3 explanatory variables .
let @xmath4 we assume that there exists @xmath5 such that @xmath6 the parameter @xmath7 and the function @xmath8 are unknown .
given @xmath2 and an unobserved error term @xmath9 the variable @xmath1 has a poisson law of mean @xmath10 if @xmath11 and @xmath12 then @xmath13 .
\label{mom2}\end{aligned}\ ] ] this model is a semiparametric single - index regression model ( e.g. , powell , stock and stoker ( 1989 ) , ichimura ( 1993 ) , hrdle , hall and ichimura ( 1993 ) , sherman ( 1994b ) ) where a second order conditional moment is specified as a nonlinear function of the conditional mean and an additional unknown parameter .
this extends the framework of cui , hrdle and zhu ( 2011 ) where the conditional variance of the response is proportional to a given function of the conditional mean .
our first contribution is to propose a new semiparametric estimation procedure for single - index regression which incorporates the additional information on the conditional variance of @xmath1 . for this
we extend the quasi - generalized pseudo maximum likelihood method introduced by gouriroux , monfort and trognon ( 1984a , 1984b ) to a semiparametric framework .
more precisely , we propose to estimate @xmath7 and the function @xmath14 through a two - step pseudo - maximum likelihood ( pml ) procedure based on _ linear exponential families _ _ with nuisance parameter _ densities .
such densities are parameterized by the mean @xmath8 and a nuisance parameter that can be recovered from the variance .
although we use a likelihood type criterion , no conditional distribution assumption on @xmath1 given @xmath2 is required for deriving the asymptotic results . as an example of application of our procedure
consider the case where @xmath1 is a count variable .
first , write the poisson likelihood where the function @xmath14 is replaced by a kernel estimator and maximize this likelihood with respect to @xmath15 to obtain a semiparametric pml estimator of @xmath7 . use this estimate and the variance formula ( [ mom2 ] ) to deduce a consistent moment estimator of @xmath16 in a second step , estimate @xmath7 through a semiparametric negative binomial pml where @xmath8 is again replaced by a kernel estimator and the variance parameter of the negative binomial is set equal to the estimate of @xmath16 finally , given the second step estimate of @xmath7 , build a kernel estimator for the regression @xmath14 . for simplicity
, we use a nadaraya - watson estimator to estimate @xmath14 .
other smoothers like local polynomials could be used at the expense of more intricate technical arguments .
the occurrence of a nonparametric estimator in a pseudo - likelihood criterion requires a rule for the smoothing parameter . while the semiparametric index regression literature contains a large amount of contributions on how to estimate an index , there are much less results and practical solutions on the choice of the smoothing parameter . even if the smoothing parameter does not influence the asymptotic variance of a semiparametric estimator of @xmath17 , in practice
the estimate of @xmath17 and of the regression function may be sensitive to the choice of the smoothing parameter .
another contribution of this paper is to propose an automatic and natural choice of the smoothing parameter used to define the semiparametric estimator . for this
, we extend the approach introduced by hrdle , hall and ichimura ( 1993 ) ( see also xia and li ( 1999 ) , xia , tong and li ( 1999 ) and delecroix , hristache and patilea ( 2006 ) ) .
the idea is to maximize the pseudo - likelihood simultaneously in @xmath15 and the smoothing parameter , that is the bandwidth of the kernel estimator .
the bandwidth is allowed to belong to a large range between @xmath18 and @xmath19 . in some sense
, this approach considers the bandwidth an auxiliary parameter for which the pseudo - likelihood may provide an estimate . using a suitable decomposition of the pseudo - log - likelihood
we show that such a joint maximization is asymptotically equivalent to separate maximization of a purely parametric ( nonlinear ) term with respect to @xmath15 and minimization of a weighted ( mean - squared ) cross - validation function with respect to the bandwidth .
the weights of this cross - validation function are given by the second order derivatives of the pseudo - log - likelihood with respect to @xmath8 .
we show that the rate of our ` optimal ' bandwidth is @xmath20 , as expected for twice differentiable regression functions .
the paper is organized as follows . in section [ metoda ]
we introduce a class of semiparametric pml estimators based on linear exponential densities with nuisance parameter and we provide a natural bandwidth choice .
moreover , we present the general methodology used for the asymptotics .
section [ rezultat ] contains the asymptotic results .
a bound for the variance of our semiparametric pml estimators is also derived . in section [ twostep ]
we use the semiparametric pml estimators to define a two - step procedure that can be applied in single - index regression models where an additional variance condition like ( [ mom2 ] ) is specified .
section [ simulsec ] examines the finite - sample properties of our procedure via monte carlo simulations .
we compare the performances of a two - step generalized least - squares with those of a negative binomial pml in a poisson single - index regression model with multiplicative unobserved heterogeneity . even if the two procedures considered lead to asymptotically equivalent estimates , the latter procedure seems preferable in finite samples .
an application to real data on the frequency of recreational trips ( see cameron and trivedi ( 2013 ) , page 246 ) is also provided .
section [ concl ] concludes the paper .
the technical proofs are postponed to the appendix .
[ metoda ] consider that the observations @xmath21 are independent copies of the random vector @xmath22 assume that there exists @xmath23 , unique up to a scale normalization factor , such that the single - index model ( sim ) condition @xmath24 holds . in this paper , we focus on single - index models where the conditional second order moment of @xmath1 given @xmath2 is a known function of @xmath25 $ ] and of a nuisance parameter .
to be more precise , in the model we consider , @xmath26 for some real value @xmath27 .
the function @xmath28 is known and , for each @xmath29 the map @xmath30 is one - to - one .
our framework is slightly more general that the one considered by cui , hrdle and zhu ( 2011 ) where the conditional variance of @xmath1 given @xmath2 is a given function of the conditional mean of @xmath1 given @xmath2 multiplied by an unknown constant . to estimate the parameter of interest @xmath7 in a model like ( [ sim])-([var1 ] )
, we propose a semiparametric pml procedure based on linear exponential families with nuisance parameter . the density used to build the pseudo - likelihood
is taken with mean and variance equal to @xmath8 and @xmath31 , respectively . in this section
we suppose that an estimator of the nuisance parameter is given . in section [ twostep ]
we show how to build such an estimator using a preliminary estimate of @xmath7 and condition ( [ var1 ] ) .
gouriroux , monfort and trognon ( 1984a ) introduced a class of densities , with respect to a given measure @xmath32 , called linear exponential family with nuisance parameter ( lefn ) and defined as @xmath33 , \ ] ] where @xmath34 is the nuisance parameter . since the dominating measure @xmath32 need not be lebesgue measure , the law defined by @xmath35 is not necessarily continuous .
the functions @xmath36 and @xmath37 are such that the expectation of the corresponding law is @xmath38 while the variance is @xmath39 ^{-1}.$ ] ( @xmath40 denotes the derivative with respect to the argument @xmath41 ) recall that for any given @xmath42 the following identity holds : @xmath43 if @xmath34 is fixed , a lefn becomes a linear exponential family ( lef ) of densities .
gouriroux , monfort and trognon ( 1984a , 1984b ) used lefn densities to define a two - step pml procedure in nonlinear regression models where a specification of the conditional variance is given . herein , we extend their approach to a semiparametric framework . in the case of the sim defined by equation ( [ mom2 ] ) , the conditional variance is given by @xmath44 with @xmath8 and @xmath45 in this case take @xmath46 which define a negative binomial distribution of mean @xmath8 and variance @xmath47 .
note that the limit case @xmath48 corresponds to a poisson distribution . as another example ,
consider @xmath49 with @xmath8 and @xmath45 now , take the lefn density given by @xmath50 and @xmath51 which is the density of a gamma law of mean @xmath38 and variance @xmath52 . in order to define our semiparametric pml estimator in the presence of a nuisance parameter
let us introduce some notation : given @xmath53 a sequence of numbers growing slowly to infinity ( e.g. , @xmath54 ) , let @xmath55 be the range from which the ` optimal ' bandwidth will be chosen .
define the set @xmath56 , @xmath57 with @xmath58 some sequence decreasing to zero .
let @xmath59 be some real value of the nuisance parameter@xmath60 typically , @xmath61 if the conditional variance formula ( [ var1 ] ) is correctly specified . otherwise , @xmath59 is some pseudo - true value of the nuisance parameter .
suppose that a sequence @xmath62 such that @xmath63 , in probability , is given .
set where @xmath64 $ ] is a lefn density .
however , other functions @xmath65 having the required properties can be considered ( see appendix [ assu ] ) . ]
@xmath66with @xmath67 the lefn density of expectation @xmath8 and nuisance parameter @xmath34 .
define the semiparametric pml estimator in the presence of a nuisance parameter and the optimal bandwidth as @xmath68 where @xmath69 denotes the leave - one - out version of the nadaraya - watson estimator of the regression function @xmath70 with @xmath71 the density of @xmath72 the function @xmath73 is a second order kernel function and @xmath74 stands for @xmath75 where @xmath76 is the bandwidth .
@xmath77 denotes a trimming function .
if the sequence @xmath78 is constant or @xmath79 does not depend on @xmath42 equation ( [ deeff ] ) defines a semiparametric pml based on a lef density .
a trimming is designed to keep the density estimator @xmath80 away from zero in computations and it is usually required for analyzing the asymptotic properties of the nonparametric regression estimator and of the ` optimal ' bandwidth .
the practical purpose of a trimming recommends a data - driven device like @xmath81 , with some fixed @xmath82 .
( herein , @xmath83 denotes the indicator function of the set @xmath84 ) however , to ensure consistency with such a trimming , one should require in addition that @xmath85 .\ ] ] meanwhile , a trimming like @xmath86 is easier to handle in theory . here
, we consider @xmath87 with @xmath88 @xmath89 , a sequence with limit @xmath7 and @xmath90 @xmath89 , a sequence of preliminary bandwidths such that @xmath91 and @xmath92 for some @xmath93 the trimming procedure we propose represents an appealing compromise between the theory and the applications . on one hand , it is easy to implement . on the other hand , we show below that , in a certain sense , our trimming is asymptotically equivalent to the fixed trimming @xmath94 and this fact greatly simplifies the proofs .
we prove this equivalence under two types of assumptions : either i ) @xmath2 is bounded and @xmath95 , or ii ) @xmath96 < \infty , $ ] for some @xmath97 and @xmath98 to be more precise , define @xmath99 and @xmath100 @xmath101 by little algebra , for all @xmath102 @xmath76 and @xmath103 @xmath104 where @xmath105 let @xmath106 with @xmath107 or @xmath108 without loss of generality , consider that @xmath109 ( since @xmath79 is the logarithm of a lefn density@xmath110 for any given @xmath111 and @xmath34 , the map @xmath112 attains its maximum at @xmath113 thus , up to a translation with a function depending only on @xmath111 and @xmath114 we may consider @xmath115 ) in this case we have @xmath116 we show that @xmath117 uniformly over @xmath118 and uniformly in @xmath42 provided that @xmath119 and @xmath120 on the other hand , we prove that @xmath121 provided that @xmath119 slowly enough and @xmath122 faster than @xmath123 and slower than @xmath124for some @xmath125 .
( see lemma [ idic ] in the appendix ; in that lemma we distinguish two types of assumptions depending on whether @xmath2 is bounded or not . )
deduce that @xmath126 is asymptotically equivalent to the maximizer of @xmath127 over @xmath128 therefore , hereafter , we simply write @xmath129 instead of @xmath130 and we consider @xmath131 the semiparametric pseudo - log - likelihood @xmath132 can be split into a purely parametric ( nonlinear ) part @xmath133 , a purely nonparametric one @xmath134 and a reminder term @xmath135 , where @xmath136 i_{a}\left ( z_{i}\right ) , \label{deco } \\ t\left ( h;\alpha ^{\ast } \right ) & = \dfrac{1}{n}\sum\limits_{i=1}^{n}\psi \left ( y_{i},\hat{r}_{h}^{i}\left ( z_{i}^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha ^{\ast } \right ) i_{a}\left ( z_{i}\right ) , \notag \\ r\left ( \theta , h;\widetilde{\alpha } _ { n}\right ) & = \dfrac{1}{n } \sum\limits_{i=1}^{n}\left [ \psi \left ( y_{i},\hat{r}_{h}^{i}\left ( z_{i}^{t}\theta ; \theta \right ) ; \widetilde{\alpha } _ { n}\right ) -\psi \left ( y_{i},r\left ( z_{i}^{t}\theta ;
\theta \right ) ; \widetilde{\alpha } _ { n}\right ) \right ] i_{a}\left ( z_{i}\right ) \notag \\
& \hspace*{0.5 cm } -\dfrac{1}{n}\sum\limits_{i=1}^{n}\left [ \psi \left ( y_{i},\hat{r } _ { h}^{i}\left ( z_{i}^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha ^{\ast}\right ) -\psi \left ( y_{i},r\left ( z_{i}^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha ^{\ast } \right ) \right ] i_{a}\left ( z_{i}\right ) \notag\end{aligned}\ ] ] ( see hrdle , hall and ichimura ( 1993 ) for a slightly different splitting ) . given this decomposition ,
the simultaneous optimization of @xmath129 is asymptotically equivalent to separately maximizing @xmath133 with respect to @xmath15 and @xmath137 with respect to @xmath76 , provided that @xmath138 is sufficiently small .
a key ingredient for proving that @xmath139 is negligible with respect to @xmath140 and @xmath141 uniformly in @xmath142 and for any @xmath143 is represented by the orthogonality conditions @xmath144 = 0 \label{orth1}\ ] ] and @xmath145 = 0,\ ] ] that must hold for any @xmath42 where @xmath146 denotes the derivative with respect to the second argument of @xmath147 and @xmath148 is the derivative with respect to all occurrences of @xmath149 that is given @xmath111 , @xmath150 and @xmath42@xmath151 ( see also sherman ( 1994b ) and delecroix , hristache and patilea ( 2006 ) for similar conditions ) . if @xmath152 with @xmath153 then @xmath154 and thus ( [ orth1 ] ) is a consequence of the sim condition ( [ sim ] ) . to check the second orthogonality condition note that @xmath155 = e\left [ \partial _ { 22}^{2}\psi \left ( y,\;r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) \mid z^{t}\theta _ { 0}\right]\ ] ] and @xmath156 = e\left [ r^{\prime } \left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \left ( z - e\left [ z\mid z^{t}\theta _ { 0}\right ] \right ) \mid z^{t}\theta _ { 0}\right ] , \ ] ] where @xmath157 is the derivative of @xmath158 the last identity is always true under the sim condition ( e.g. , newey ( 1994 ) , page 1358)@xmath60 let us point out that conditions ( [ orth1])-([orth2 ] ) hold even if the variance condition ( [ var1 ] ) is misspecified . since @xmath138 is negligible with respect to @xmath159 and @xmath137
does not contain the parameter of interest , the asymptotic distribution of @xmath160 will be obtained by standard arguments used for @xmath161estimators in the presence of nuisance parameters applied to the objective function @xmath133 .
we deduce that @xmath160 behaves as follows :
i ) if the sim condition ( [ sim ] ) holds and @xmath162 for some @xmath163 then @xmath160 is asymptotically normal ; ii ) if sim condition holds , the conditional variance ( [ var1 ] ) is correctly specified and @xmath164 then @xmath160 is asymptotically normal and it has the lowest variance among the semiparametric pml estimators based on lef densities . in any case , the asymptotic distribution of @xmath165 does not depend on the choice of @xmath166 let us point out that in our framework we only impose @xmath78 convergent in probability without asking a rate of convergence @xmath167 as it is usually supposed for @xmath161estimation in the presence of nuisance parameters .
this because the usual orthogonality condition @xmath168 = 0 $ ] is true for any @xmath42 provided that @xmath169 with @xmath67 a lefn density@xmath60 indeed , we have @xmath170 \\ & = e\left [ \partial _ { \alpha } \partial _ { r}\psi \left ( y , r\left ( z^{t}\theta_{0};\theta _ { 0}\right ) ; \alpha \right ) \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \right ] \\ & = e\left [ e\left\ { \partial _ { \alpha } \partial _ { r } b\left ( r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) + \partial
_ { \alpha } \partial _ { r } c\left ( r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) y\mid z\right\ } \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \right ] \\ & = 0\end{aligned}\ ] ] because @xmath171 and @xmath172 for any @xmath173 for the bandwidth @xmath174 we obtain an asymptotic equivalence with a theoretical ` optimal ' bandwidth minimizing @xmath175 that is we prove that the ratio of the two bandwidths converges to one , in probability .
remark that @xmath176 is a kind of @xmath177cv ( cross validation ) function .
it can be shown that , up to constant additive terms , @xmath176 is asymptotically equivalent to a weighted ( mean - squared ) cv function .
when @xmath178 the function @xmath176 is the usual cv function that one would use for choosing the bandwidth for the nadaraya - watson estimator of @xmath179
. by extension of classical results for nonparametric regression , it can be proved that the rate of the theoretical ` optimal ' bandwidth minimizing @xmath176 is @xmath20 ( see lemma [ mise ] in appendix [ appconv ] ; see also hrdle , hall and ichimura ( 1993 ) for the case @xmath180)@xmath60 deduce that @xmath174 is also of order @xmath181 given the model conditions ( [ sim])-([var1 ] ) , the idea is to choose a lefn density with mean @xmath8 and variance @xmath31 and to construct a semiparametric pml estimator given a preliminary estimate of the nuisance parameter @xmath27 .
however , it may happen that no such lefn density exists or that one prefers another type of lefn densities .
then , the idea is to reparametrize the conditional variance of @xmath1 given @xmath2 .
more precisely , we may consider @xmath182 , \ ] ] where @xmath183 stands for the nuisance parameter .
let @xmath184 denote the variance of the law given by this density .
assume that for any given @xmath8 , the map @xmath185 is one - to - one .
in this case , in order to provide a lefn density with variance @xmath31 it suffices to consider @xmath186 with @xmath187 .
for instance , if @xmath188 , one may use a negative binomial density of mean @xmath8 and nuisance parameter @xmath189 .
another solution is to consider a normal density of mean @xmath8 where the variance equal to @xmath190 plays the role of the nuisance parameter . in this case ,
given an estimate of @xmath190 , our semiparametric pml becomes a semiparametric generalized least - squares ( gls ) procedure .
note that this example of function @xmath191 leads us to the situation where the nuisance parameter is replaced by a ` nuisance ' function of @xmath8 and some additional parameters . at the expense of more complicated writings
, our methodology can be extended to take into account the case of a ` nuisance ' function .
more precisely , consider a more general pseudo - log - likelihood function @xmath192 where @xmath193 is a given real - valued function and @xmath34 is the nuisance parameter .
see also gouriroux , monfort and trognon ( 1984a ) . to define @xmath194 one
replaces @xmath78 by @xmath195 in equation ( [ deeff ] ) , where @xmath196 , in probability , for some @xmath197 and @xmath198 is a nadaraya - watson estimator of the regression @xmath199 the same type of decomposition of the pseudo - log - likelihood criterion into a purely parametric part function of @xmath15 @xmath200 i_{a}\left ( z_{i}\right ) , \end{gathered}\ ] ] a purely nonparametric part function of @xmath76 @xmath201 and a negligible reminder function of @xmath15 and @xmath76 can be used . for brevity , the details of this more general case are omitted .
however , we sketch a quick argument that applies for the semiparametric gls .
consider the semiparametric gls criterion @xmath202 ^{2}i_{a}\left ( z_{i}\right)\ ] ] with @xmath203 , in probability , and @xmath90 @xmath57 a sequence of bandwidths .
assume that @xmath204 and @xmath205 stays away from zero .
then the gls criterion @xmath206 is asymptotically equivalent to the infeasible gls criterion @xmath207 ^{2}g\left ( r(z_{i}^{t}\theta _ { 0};\theta _ { 0});\alpha ^{\ast } \right)^{-1}i_{a } \left ( z_{i}\right ) , \ ] ] that is we can decompose the two criteria in such way that , up to negligible reminders , they have exactly the same purely parametric and purely nonparametric parts .
finally , we apply the methodology with @xmath208 can be written as a function of @xmath209 in view of our proofs , it becomes obvious that the methodology described in the previous subsection remains valid if @xmath210 is multiplied by a function depending only on @xmath211 described in the previous subsection with @xmath180 and the trimming @xmath210 multiplied by @xmath212 in order to ensure condition ( [ uunif ] ) , it suffices to suppose that the map @xmath213 satisfies a lipschitz condition and that @xmath214 is such that @xmath215and @xmath216 is bounded in probability , uniformly with respect to @xmath15 in @xmath217 neighborhoods of @xmath218 for instance , a bandwidth of order @xmath20 satisfies these conditions ( see andrews ( 1995 ) ; see also delecroix , hristache and patilea ( 2006 ) ) .
other possible extensions of the framework we consider is to allow a multi - index regression and/or multivariate dependent variables .
for instance , the sim condition can be replaced by the multi - index condition @xmath219 with @xmath220 smaller than the dimension of @xmath221 while the second order moment condition remains @xmath222 on the other hand , for multivariate dependent variables one may consider pml estimation based on the multivariate normal or multivariate generalizations of poisson , negative binomial distributions ( johnson , kotz and balakrishnan ( 1997 ) ) .
the decomposition of the pseudo - log - likelihood in @xmath223 , @xmath224 and @xmath225 as above can still be used for these cases but the detailed analysis of these extensions will be considered elsewhere .
in this section we obtain the asymptotic distribution for @xmath160 and the corresponding estimator of the regression function @xmath226 $ ] as well as the asymptotic behavior of @xmath174 , with @xmath227 defined in ( [ deeff ] ) .
a consistent estimator for the asymptotic variance matrix of @xmath160 is proposed@xmath60 moreover , a lower bound for the asymptotic variance matrix of @xmath160 is derived . for the identifiability of the parameter of interest @xmath7 , hereafter fix its first component ,
that is @xmath228 @xmath229 therefore , we shall implicitly identify a vector @xmath230 with its last @xmath231 components and redefine the symbol @xmath148 as being the vector of the first order partial derivatives with respect to the last @xmath231 components of @xmath232 let @xmath233 if the sim assumption and variance condition ( [ var1 ] ) hold , then @xmath234 for a given @xmath149 let @xmath235 and @xmath236 denote the first and second order derivatives of the function @xmath237 similarly , @xmath238 is the derivative of @xmath239 define thus , @xmath240 can be replaced by @xmath241 in the definition of the constants @xmath242 and @xmath243 @xmath244 ^{2}\ i_{a}\left ( z\right ) \right\ } \notag \\ c_{2 } & = -\,k_{2}\ e\left\ { \frac{1}{2}\ \partial _ { r}c\left ( r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha ^{\ast } \right ) \,\,\frac{1}{f\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) } \ v\left ( z^{t}\theta_{0};\theta _ { 0}\right ) \ i_{a}\left ( z\right ) \right\ } , \notag\end{aligned}\ ] ] with @xmath245 , @xmath246 and consider @xmath247define the @xmath248 matrices @xmath249 ^{2}v\left ( z^{t}\theta_{0};\theta _ { 0}\right ) \partial _ { \theta } r\left ( z^{t } \theta _ { 0};\theta_{0}\right ) \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ^{t}i_{a}\left ( z\right ) \right\}\ ] ] @xmath250 .\ ] ] note that @xmath251 if the variance condition ( [ var1 ] ) holds and @xmath252 now , we deduce the asymptotic normality of the semiparametric pml @xmath160 estimator in the presence of a nuisance parameter .
moreover , we obtain the rate of decay to zero of the ` optimal ' bandwidth @xmath253 the proof of the following result is given in appendix refproof .
[ param1 ] suppose that the assumptions in appendix [ assu ] hold .
define the set @xmath254 , @xmath89 , with @xmath255 and @xmath256 @xmath57 such that @xmath257 .
fix @xmath258 if @xmath227 is defined as in ( [ deeff])-([deeff2 ] ) , then @xmath259 in probability , and @xmath260 if @xmath2 is bounded , the same conclusion remains true for any sequence @xmath261 . in applications
@xmath262 is unknown and therefore it has to be consistently estimated . to this end
, we propose an usual sandwich estimator of the asymptotic variance @xmath262 ( e.g. , ichimura ( 1993 ) ) .
let @xmath263 denote the kernel estimator for the density of @xmath72 define @xmath264 ^{2}\left [ y_{i}-\widehat{r}_{\widehat{h}}\left ( z_{i}^{t}\widehat{\theta } ; \widehat{\theta } \right ) \right ] ^{2 } \\
_ { \theta } \widehat{r}_{\widehat{h}}\left ( z_{i}^{t}\widehat{\theta } ; \widehat{\theta } \right ) \partial _ { \theta } \widehat{r}_{\widehat{h } } \left ( z_{i}^{t } \widehat{\theta } ; \widehat{\theta } \right ) ^{t}i_{\left\{z:\,\widehat{f}_{\widehat{h}}\left ( z^{t}\widehat{\theta } ; \,\widehat{\theta } \right ) \geq c\right\ } } ( z_{i})\end{gathered}\ ] ] @xmath265 [ estasv ] suppose that the conditions of theorem [ param1 ] hold .
then , @xmath266 in probability .
the arguments are quite standard ( e.g. , ichimura ( 1993 ) , section 7 ) . on one hand , the convergence in probability of @xmath160 and @xmath78 and , on the other hand , the convergence in probability of @xmath267 and @xmath268 uniformly over @xmath15 in neighborhoods shrinking to @xmath7 and uniformly over @xmath269 ( e.g. , andrews ( 1995 ) , delecroix , hristache and patilea ( 2006 ) ) imply @xmath270 and @xmath271 in probability .
theorem [ param1 ] shows , in particular , that @xmath160 is asymptotically equivalent to the semiparametric pml based on the lef pseudo - log - likelihood @xmath272 as in the parametric case , we can deduce a lower bound for the asymptotic variance @xmath262 with respect to semiparametric pml based on lef densities .
this bound is achieved by @xmath160 if the sim assumption and the variance condition ( [ var1 ] ) hold and @xmath273 the proof of the following proposition is identical to the proof of property 5 of gouriroux , monfort and trognon ( 1984a , page 687 ) and thus it will be skipped .
[ bound ] the set of asymptotic variance matrices of the semiparametric pml estimators based on linear exponential families has a lower bound equal to @xmath274 , where @xmath275 ^{-1}\partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \partial _ { \theta } r\left ( z^{t}\theta_{0 } ; \theta _ { 0}\right ) ^{t}i_{a}\left ( z\right ) \right\ } .\ ] ] concerning the nonparametric part , we have the following result on theasymptotic distribution of the nonparametric estimator of the regression .
the proof is omitted ( see hrdle and stoker ( 1989 ) ) .
[ nonpar ] assume that the conditions of theorem [ param1 ] are fulfilled .
then , for any @xmath276 such that @xmath277 @xmath278 where @xmath279 .$ ] note that , for any @xmath150 such that @xmath280 @xmath281 indeed , use the results of andrews ( 1995 ) to deduce that @xmath282 in probability , uniformly over neighborhoods of @xmath7 where @xmath283 stays away from zero .
therefore , we can write @xmath284 and obtain the asymptotic normality of @xmath285 as a consequence of the @xmath286consistency of @xmath160 and the asymptotic behavior of the nadaraya - watson estimator .
here , we consider a two - step semiparametric pml procedure that can be applied in semiparametric single - index regression models when a conditional variance condition like @xmath287 is specified .
assume that this conditional variance condition is correctly specified .
at the end of this section we also discuss the misspecification case .
first , we have to build a sequence @xmath288 with limit @xmath218 moreover , in the case of unbounded covariates , @xmath289 should approach @xmath7 faster than @xmath290 for this purpose , we maximize with respect to @xmath15 a pseudo - likelihood based on a lef density @xmath291 .
we use a fixed trimming @xmath292 with @xmath293 a subset of @xmath294 such that , for any @xmath15 and any @xmath295 we have @xmath296 to ensure consistency for such a pml estimator , we have to check that @xmath297 , \label{idfg}\ ] ] and @xmath7 is unique with this property . recall that the sim condition specifies @xmath7 as the unique vector satisfying @xmath298 = e\left [ y\mid z^{t}\theta _ { 0}\right ] .$ ] on the other hand , if @xmath299 then @xmath300 ( cf . property 4 , gouriroux , monfort and trognon ( 1984a , page 684 ) ) .
deduce that for any @xmath150 , @xmath301\ ] ] and @xmath7 is the unique maximizer .
hence , condition ( [ idfg ] ) holds for any set @xmath302 this leads us to the following definition of a preliminary estimator
. * step 1 ( preliminary step ) . *
consider a sequence of bandwidths @xmath303 @xmath57 such that @xmath304 and @xmath92 for some @xmath93 moreover , let @xmath291 be a lef density .
delecroix , hristache and patilea ( 2006 ) showed that , under the regularity conditions required by theorem [ param1 ] , we have @xmath306 using the preliminary estimate @xmath289 and the variance condition ( [ gmm ] ) we can build @xmath256 @xmath57 such that @xmath307 , in probability ( see the end of this section ) .
let @xmath67 denote a lefn density with mean @xmath8 and variance @xmath308 consider @xmath309 ( e.g. , @xmath54 ) , define @xmath310 .
moreover , consider @xmath311 , @xmath89 with @xmath58 as in theorem [ param1 ] .
fix some small @xmath258 * step 2 . *
define @xmath312 with @xmath313 and @xmath314 from step 1 .
the following result is a direct consequence of theorem [ param1 ] .
[ 2step ] suppose that the assumptions of theorem [ param1 ] hold .
if @xmath160 and @xmath174 are obtained as in step 2 above , then @xmath315 with @xmath275 ^{-1}\partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ^{t}i_{a}\left ( z\right ) \right\ } .\ ] ] moreover , @xmath316 in probability , where @xmath242 and @xmath317 are defined as in ( [ a1a2 ] ) with @xmath318 _ remark 1 .
_ let us point out that simultaneous optimization of the semiparametric criterion in step 1 with respect to @xmath15 , @xmath34 and @xmath76 ( or with respect to @xmath15 and @xmath34 for a given @xmath76 ) is not recommended , even if the conditional variance @xmath319 is correctly specified .
indeed , if the true conditional distribution of @xmath1 given @xmath2 is not the one given by the lefn density @xmath320 joint optimization with respect to @xmath15 and @xmath34 leads , in general , to an inconsistent estimate of @xmath321 ( this failure is well - known in the parametric case where @xmath8 is a known function ; see comments of cameron and trivedi ( 2013 ) , pages 84 - 85 . in view of decomposition ( [ deco ] ) we deduce that this fact also happens in the semiparametric framework where @xmath8 has to be estimated . ) in this case the matrices @xmath322 and @xmath323 defined in section [ rezultat ] are no longer equal and thus the asymptotic variance of the one - step semiparametric estimator of @xmath15 obtained by simultaneous maximization of the criterion in step 1 with respect to @xmath15 , @xmath34 does not achieve the bound @xmath274 .
however , when the sim condition holds and the true conditional law of @xmath1 is given by the lefn density @xmath324 our two - step estimator @xmath160 and the semiparametric mle of @xmath7 obtained by simultaneous optimization with respect to @xmath325 are asymptotically equivalent .
_ remark 2 .
_ note that if we ignore the efficiency loss due to trimming , @xmath274 is equal to the efficiency bound in the semiparametric model defined _ only _ by the single - index condition @xmath326 when the variance condition ( [ gmm ] ) holds . to see this ,
apply the bound of newey and stoker ( 1993 ) with the true variance given by ( [ gmm ] ) .
our two - stage estimator achieves this sim efficiency bound ( if the variance is well - specified ) .
however , this sim bound is not necessarily the two moment conditions model bound .
the latter should take into account the variance condition ( see newey ( 1993 ) , section 3.2 , for a similar discussion in the parametric nonlinear regression framework ) .
in other words our two - stage estimator has some optimality properties but it may not achieve the semiparametric efficiency bound of the two moment conditions model .
the same remark applies for the two - stage semiparametric generalized least squares ( gls ) procedure of hrdle , hall and ichimura ( 1993 ) [ see also picone and butler ( 2000 ) ] . achieving semiparametric efficiency when the first two moments are specified would be possible , for instance , by estimating higher orders conditional moments nonparametrically .
however , in this case we face again the problem of the curse of dimensionality that we tried to avoid by assuming the sim condition .
to complete the definition of the two - step procedure above , we have to indicate how to build a consistent sequence @xmath62
. such a sequence can be obtained from the moment condition ( [ gmm ] ) after replacing @xmath327 by a suitable estimator .
this kind of procedure is commonly used in the semiparametric literature ( e.g. , newey and mcfadden ( 1994 ) ) . for simplicity ,
let us only consider the negative binomial case where , for any @xmath328 we have @xmath329 = r\left ( z^{t}\theta_{0 } ; \theta _ { 0}\right ) \left [ 1+\alpha _ { 0}r\left ( z^{t}\theta_{0};\theta _ { 0}\right ) \right ] .\ ] ] consider a set @xmath330 such that , for any @xmath15 and any @xmath295 we have @xmath296 we can write @xmath331 i_{b}\left ( z\right ) \right\ } = \alpha_{0 } e\left\ { r\left ( z^{t}\theta_{0};\theta _ { 0}\right ) ^{2}i_{b}\left ( z\right ) \right\ } .\ ] ] consequently , we may estimate .
this is indeed confirmed by the simulation experiments we report in section [ simulsec ] . ]
@xmath27 by @xmath332 i_{b}\left ( z_{i}\right ) } { \frac{1}{n } \sum_{i=1}^{n}\widehat{r}_{h_{n}}\left ( z_{i}^{t}\theta _ { n};\theta _ { n}\right ) ^{2}i_{b}\left ( z_{i}\right ) } \label{nuis}\]]with @xmath289 and @xmath214 from step 1 and @xmath333 the nadaraya - watson estimator with bandwidth @xmath214 .
since @xmath334 deduce that @xmath335 in probability ( see also the arguments we used in subsection [ relax ] ) .
now , let us comment on what happens with our two - step procedure if the second order moment condition is misspecified , while the sim condition still holds .
in general , the sequence @xmath78 one may derive from the conditional variance condition and the preliminary estimate of @xmath7 is still convergent to some pseudo - true value @xmath59 of the nuisance parameter . defined in ( [ nuis ] ) is convergent in probability to @xmath336-e[r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) i_{b}(z)]}{e[r(z^{t}\theta _ { 0};\theta _ { 0})^{2 } i_{b}(z)]}.\ ] ] to ensure that the limit of @xmath78 is positive , one may replace @xmath78 by @xmath337 for some small but positive @xmath338 then , the behavior of @xmath126 yielded by step 2 is described by theorem [ param1 ] , that is @xmath160 is still @xmath286asymptotically normal and @xmath174 is still of order @xmath181 finally , if the sim condition does not hold , then @xmath160 estimates a kind of first projection - pursuit direction . in this case
, our procedure provides an alternative to minimum average ( conditional ) variance estimation ( mave ) procedure of xia _ et al_. ( 2002 ) .
the novelty would be that the first projection direction is defined through a more flexible pml function than the usual least - squares criterion .
this case will be analyzed elsewhere .
in our empirical section we consider the case of a count response variable @xmath1 .
a benchmark model for studying event counts is the poisson regression model .
different variants of the poisson regression have been used in applications on the number of patents applied for and received by firms , bank failures , worker absenteeism , airline or car accidents , doctor visits , _ etc . _
cameron and trivedi ( 2013 ) provide an overview of the applications of poisson regression . in the basic setup ,
the regression function is log - linear .
an additional unobserved multiplicative random error term in the conditional mean function is usually used to account for unobserved heterogeneity . in this section
we consider semiparametric single - index extensions of such models . to evaluate the finite sample performances of our estimator @xmath339 and of the optimal bandwidth @xmath174 , we conduct a simulation experiment with 500 replications .
we consider three explanatory variables @xmath340 with @xmath341_{3\times3}$ ] and @xmath342 .
the regression function is @xmath343 and @xmath344 .
the conditional distribution of @xmath1 given @xmath2 and @xmath345 is poisson of mean @xmath346 with @xmath345 independent of @xmath2 and distributed according to @xmath347 or @xmath348 .
thus , the conditional variance of @xmath1 given @xmath2 is given by the function @xmath44 with @xmath349 for @xmath350 and @xmath351 for @xmath352 .
for this simulation experiment we generate samples of size @xmath353 and @xmath354 . for the nonparametric part we use a quartic kernel @xmath355 } \left ( u\right ) $ ] . to estimate the parameter @xmath7 and the regression @xmath356 we use two semiparametric two - step estimation procedures as defined in section [ twostep ] : i ) a procedure with a poisson pml in the first step and a negative binomial pml in the second step ; let @xmath357 denote the two - step estimator .
ii ) a procedure with a least - squares method in the first step and a gls method in the second step ; let @xmath358 be the two - step estimator . note that @xmath359 and @xmath360 have the same asymptotic variance . in both two - step procedures considered , we estimate @xmath27 using the estimator defined in ( [ nuis ] ) .
the bandwidth @xmath214 is equal to @xmath361 we also consider the parametric two - step gls method as a benchmark . in this case
the link function and the variance parameter are considered given ;
let @xmath362 denote the corresponding estimator .
table 2 . the same setup as in table 1 but with @xmath352 and the true conditional variance of @xmath1 given @xmath2 equal to @xmath363 [ cols="^,^,^,^,^,^,^,^,^,^ " , ]
we consider a semiparametric single - index model ( sim ) where an additional second order moment condition is specified . to estimate the parameter of interest @xmath15 we introduce a two - step semiparametric pseudo - maximum likelihood ( pml ) estimation procedure based on linear exponential families with nuisance parameter densities .
this procedure extends the quasi - generalized pseudo - maximum likelihood method proposed by gouriroux , monfort and trognon ( 1984a , 1984b ) .
we also provide a natural rule for choosing the bandwidth of the nonparametric smoother appearing in the estimation procedure .
the idea is to maximize the pseudo - likelihood of the second step simultaneously in @xmath15 and the smoothing parameter @xmath76 .
the rate of the bandwidth is allowed to lie in a range between @xmath18 and @xmath19 .
we derive the asymptotic behavior of @xmath160 , the two - step semiparametric pml we propose . if the sim condition holds , then @xmath160 is @xmath286asymptotically normal .
we also provide a consistent estimator of its variance . when the sim condition holds and the conditional variance is correctly specified , then @xmath160 has the best variance amongst the semiparametric pml estimators .
the ` optimal ' bandwidth @xmath174 obtained by joint maximization of the pseudo - likelihood function in the second step is shown to be equivalent to the minimizer of a weighted cross - validation function . from this
we deduce that @xmath364 converges to a positive constant , in probability .
in particular , our optimal bandwidth @xmath174 has the rate expected when estimating a twice differentiable regression function nonparametrically .
we conduct a simulation experiment in which the data were generated using a poisson single - index regression model with multiplicative unobserved heterogeneity .
the simulation confirms the significant advantage of estimators that incorporate the information on the conditional variance .
we also applied our semiparametric approach to a benchmark real count data set and we obtain a much better fit than the standard parametric regression models for count data .
let @xmath365 with @xmath366 a compact subset of @xmath367 with nonvoid interior . depending on the context ,
@xmath368 is considered a subset of @xmath367 or a subset of @xmath369 [ asiid]the observations @xmath370 are independent copies of a random vector @xmath371 let @xmath372 there exists a unique @xmath7 interior point of @xmath368 such that @xmath373 [ as41]for every @xmath374 , the random variable @xmath375 admits a density @xmath376 with respect to the lebesgue measure on @xmath377 .
[ as42]@xmath96 < \infty , $ ] for some @xmath378 .
moreover , @xmath379 for some @xmath380 .
[ colin ] with probability one , the matrix @xmath381 is positive definite .
[ cc ] there exists @xmath382 and a positive integer @xmath383 such that , for any @xmath374 and @xmath384 , the set @xmath385 has at most @xmath383 elements .
the last two assumptions ensure that @xmath386 for any @xmath387 * condition l * a function @xmath388 is said to satisfy _
condition l _ if , for any @xmath389 a compact set on the real line , there exists @xmath390 and @xmath391 $ ] such that @xmath392 [ as45 ] a ) the function @xmath393 @xmath394 @xmath395 satisfies a lipschitz condition , that is there exists @xmath396 $ ] and @xmath397 such that @xmath398 \b ) the function @xmath399 @xmath394 @xmath395 satisfies _ _ condition l__@xmath60 \c ) for any @xmath394 the functions @xmath400 and @xmath401 are twice differentiable .
let @xmath402 and @xmath403 denote the second order derivatives . the functions @xmath404 and @xmath405 @xmath406 @xmath395 satisfy _ condition l _ with @xmath407__. _ _
\d ) for any @xmath374 and any component @xmath408 of @xmath221 the functions @xmath409 and @xmath410 are twice differentiable and their second order derivatives satisfy _ condition l _ with @xmath407__. _ _
\e ) for any @xmath395 the function @xmath411 is twice continuously differentiable and , for any @xmath412 the functions @xmath413 and @xmath414 are continuous .
moreover , the function @xmath415 satisfy _ condition l _ with @xmath407__. _
_ let @xmath416 be the conditional variance of @xmath1 given @xmath417 [ as46]the function @xmath418 satisfies _
condition l_. consider the functions @xmath419 , with @xmath420 , @xmath421 @xmath422 .
define @xmath423 , with @xmath424 , @xmath425 , and @xmath426 [ asssup]if @xmath82 , there exists @xmath425 such that @xmath427 is strictly included in @xmath225 .
[ as410]the kernel function @xmath73 is differentiable , symmetric , positive and compactly supported .
moreover , @xmath428 and the derivative @xmath429 are of bounded variation . up to a term depending only on @xmath111 and
@xmath42 the three arguments function @xmath147 involved in equation ( [ deeff ] ) is defined as @xmath430where @xmath431 $ ] is a lefn density with mean @xmath8 and variance @xmath432 ^{-1}.$ ] [ as49 ] the functions @xmath433 and @xmath434 are twice differentiable in the first argument .
moreover , for any @xmath424 and @xmath425 for which @xmath427 is strictly included in @xmath225 , there exists a constant @xmath435 such that @xmath436@xmath437where @xmath438 stands for @xmath293 or @xmath439 .
the functions @xmath440 and @xmath441 are continuously differentiable in @xmath442 [ as51]for any @xmath424 and @xmath425 for which @xmath427 is strictly included in @xmath421 we have @xmath443 @xmath444 @xmath445 assumption [ as51 ] ensures that the @xmath446 matrix@xmath447 \\ & = & e\left [ \partial _ { r}c\left ( r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha ^{\ast } \right ) \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \partial _ { \theta } r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ^{t}i_{a}\left ( z\right ) \right]\end{aligned}\]]is positive definite .
let us notice that the asymptotic results remain valid even if the function @xmath65 is not the logarithm of a lefn .
it suffices to adapt assumption [ as49 ] , to suppose that there exists @xmath448 such that @xmath449 @xmath450 to ensure that @xmath323 is positive definite and to assume that , for any @xmath451 , @xmath452 = 0 $ ] and @xmath453 = 0 .\ ] ]
let @xmath454 , $ ] with @xmath455 and @xmath456 , with @xmath457 .
we use @xmath439 to denote a positive constant , not necessarily the same at each occurrence .
[ lemrate ] assume that the kernel @xmath458 is a symmetric , positive , compactly supported function of bounded variation .
suppose that the map @xmath459 @xmath394 @xmath395 satisfies a lipschitz condition , that is there exists @xmath396 $ ] and @xmath397 such that @xmath460 then @xmath461 the proof of lemma [ lemrate ] can be distilled from many existing results ( e.g. , andrews ( 1995 ) , sherman ( 1994b ) , delecroix , hristache and patilea ( 2006 ) ) and therefore it will be omitted . [ idic ] a ) if @xmath462 then @xmath463 where @xmath464 and @xmath465 \b ) suppose that @xmath73 and @xmath466 satisfy the assumptions of lemma [ lemrate ] for some @xmath467 .
moreover , assume that either i ) @xmath2 is bounded and @xmath261 or , ii ) @xmath96 < \infty $ ] for some @xmath378 and @xmath468 ( with @xmath469 from the definition of @xmath470)@xmath60 let @xmath471 such that @xmath472 and either i ) @xmath473 if @xmath2 is bounded or , ii ) @xmath474 ^{-a}\rightarrow \infty $ ] .
then @xmath475 @xmath476 \a ) we have @xmath477 for any @xmath478 and @xmath479 we can write @xmath480 and @xmath481 which proves the inequality .
b ) it suffices to prove that @xmath482 first consider the case of unbounded @xmath483 note that , for any @xmath150 and @xmath149 @xmath484combine this inequality and lemma [ lemrate ] and write @xmath485on the other hand , we can write @xmath486 \\ & \leq & n\frac{e^{\lambda } e\left [ \exp \left ( \lambda \left\| z_{i}\right\| \right ) \right ] } { \exp \left ( \lambda \delta _ { n}^{1/a}/d_{n}\right ) } .\end{aligned}\]]since @xmath487 and @xmath488 deduce that @xmath489 if @xmath2 lies in a compact , condition ( [ lip1 ] ) implies that for any @xmath150 in the support of @xmath2 , @xmath490 with @xmath397 some constant independent of @xmath491 in this case @xmath492 thus @xmath493 provided that @xmath494 such that @xmath495 and @xmath496 the proofs of the following three lemmas are lengthy and technical .
these proofs are provided in delecroix , hristache and patilea ( 2006 ) and therefore it will be omitted herein .
the key ingredients for the three proofs are the results on uniform rates of convergence for @xmath497processes indexed by euclidean families ; see sherman ( 1994a ) .
see also pakes and pollard ( 1989 ) for the definition and the properties of euclidean families of functions .
the first of the three lemmas is a refined version of a standard result for cross - validation in nonparametric regression ( e.g. , hrdle and marron ( 1985 ) ) .
the result holds uniformly in @xmath15,@xmath34 and for @xmath76 in @xmath498 , $ ] with @xmath93 [ mise ] suppose that assumptions [ asiid ] to [ as46 ] hold .
fix some small @xmath82 and let @xmath423 .
consider a family of functions @xmath499 , @xmath394 @xmath500 for which there exist a real - valued function @xmath501 with @xmath502<\infty $ ] , for some @xmath380 , and @xmath503 $ ] such that , for each @xmath111 @xmath504 for any @xmath505 , @xmath506 , and @xmath507 .
moreover , there exist @xmath508 and @xmath509 such that @xmath510 and @xmath511<\infty $ ] . for @xmath512 and @xmath513 define @xmath514 ^{2}\!i_{\left\ { z:\,f\left ( z^{t}\theta ; \theta \right ) \geq c\right\ } } \left ( z_{i}\right ) \quad ; \]]the kernel is a continuous probability density function @xmath458 with the support in @xmath515.$ ] moreover , @xmath458 is of bounded variation and symmetric . then , @xmath516where @xmath517 ^{2}\!\!\!i_{\left\ { z : f\left ( z^{t}\theta ; \theta \right ) \geq c\right\ } } \!\left ( z\right ) \ ! \ ! \right\ } \ ! , \notag \\ c_{2}\left ( \theta ; \alpha \right ) \!\!\!\ ! & = & \!\!\ !
k_{2}\ e\left\ { -\,\,\frac{w_{\theta , \alpha } \left ( y , z^{t}\theta \right ) } { f\left ( z^{t}\theta ; \theta \right ) } \ v\left ( z^{t}\theta ; \theta \right ) \,i_{\left\ { z:\,f\left ( z\theta ; \theta \right ) \geq c\right\ } } \left ( z\right ) \right\ } \ ! , \notag\end{aligned}\]]with @xmath245 , @xmath518 and @xmath519 [ mise2 ] assume that the conditions of lemma [ mise ] hold . let @xmath520 \
i_{\left\ { z:\,f\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) \geq c\right\ } } \left ( z_{i}\right ) , \]]where , for some @xmath380 , @xmath521 = 0 $ ] and @xmath522 < \infty $ ] , for any @xmath451 . then , @xmath523uniformly in @xmath524 , in @xmath525 and in @xmath526 the following lemma will provide the order of the reminder term in the decomposition ( [ deco ] ) .
the proof relies on orthogonality conditions like ( [ orth1 ] ) and ( [ orth2 ] ) .
[ rr ] let @xmath527 for some @xmath258 let @xmath528 such that such that @xmath529 suppose that the assumptions [ asiid ] , as41 , [ as42 ] and [ as45 ] to [ as49 ] hold . then @xmath530 \ i_{a}\left ( z_{i}\right ) \\ & & -\dfrac{1}{n}\sum\limits_{i=1}^{n}\left [ \phi \left ( y_{i},\hat{r}_{h}^{i}\left ( z_{i}^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) -\phi \left ( y_{i},r\left ( z_{i}^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) \right ] \ i_{a}\left ( z_{i}\right ) \\ & & \\ & = & \left [ o_{p}\left ( h^{4}\right ) + o_{p}\left ( \frac{1}{nh^{2}}\right ) + o_{p}\left ( \frac{h^{2}}{\sqrt{n}}\right ) + o_{p}\left ( \frac{1}{n\sqrt{n}h^{4}}\right ) \right ] \times o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| \right ) \\ & & + \left [ o\left ( h^{2}\right ) + o_{p}\left ( \frac{1}{\sqrt{n}h^{2}}\right ) \right ] \times o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| ^{2}\right ) \\ & & \end{aligned}\ ] ] when @xmath531 , uniformly in @xmath42 uniformly in @xmath532 , $ ] with @xmath455 and uniformly in @xmath533
assume for the moment that @xmath126 are defined by maximization of @xmath129 which is defined with the fixed trimming @xmath534 ( see equation ( [ infeasi ] ) ) .
at the end of the proof we show that the same conclusions hold for @xmath126 defined in ( [ deeff ] ) with the data - driven trimming . _
part i : _ @xmath286__asymptotic normality of _ _ @xmath535 _ _ _ _ by the decomposition ( [ deco ] ) we have @xmath536our objective is to show that @xmath138 is negligible when compared to @xmath537 from which we deduce that @xmath538 behaves as the maximizer of @xmath539 .
define @xmath540 i_{a}\left ( z_{i}\right ) \;\]]and use taylor expansion to write @xmath541 + \left [ r_{1}\left ( \theta _ { 0},h;\widetilde{\alpha } _ { n}\right ) -r_{1}\left
( \theta _ { 0},h;\alpha ^{\ast } \right ) \right ] .\end{aligned}\]]apply lemma [ rr ] to obtain the order of @xmath542 next , note that @xmath543 does not depend on @xmath15 .
deduce that @xmath544 \times o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| \right ) \\ & & + \left [ \!o\left ( h^{2}\right ) + o_{p}\left ( \frac{1}{\sqrt{n}h^{2}}\right ) \!\right ] \times o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| ^{2}\right ) \\ & & + \left\ { \text{terms not depending on } \theta \right\ } , \end{aligned}\]]uniformly in @xmath532 , $ ] with @xmath455 uniformly in @xmath525 and uniformly with respect to @xmath545 it follows that , up to terms not containing @xmath149 @xmath546uniformly in @xmath524 , @xmath525 and uniformly with respect to @xmath545 now , write @xmath547 with @xmath548notice that @xmath549 + \left [ \widetilde{s}_{1}\left ( \theta _ { 0},\widetilde{\alpha } _ { n}\right ) -\widetilde{s}_{1}\left ( \theta _ { 0},\alpha ^{\ast } \right ) \right ] , \]]where the last difference does not contain @xmath15 , so that @xmath550.\]]furthermore , for any @xmath551 in probability , we have @xmath552 therefore , using the taylor expansion we can write @xmath553 \\ & & + \left ( \theta -\theta _ { 0}\right ) ^{t}\,\partial _ { \theta \theta ^{t}}^{2}\widetilde{s}_{1}\left ( \theta _ { 0},\alpha ^{\ast } \right ) \,\left ( \theta -\theta _ { 0}\right ) \\ & & + \left ( \theta -\theta _ { 0}\right ) ^{t}\,\left [ \partial _ { \theta \theta ^{t}}^{2}\widetilde{s}_{1}\left ( \theta _ { 0},\widetilde{\alpha } _ { n}\right ) -\partial _ { \theta \theta ^{t}}^{2}\widetilde{s}_{1}\left ( \theta _ { 0},\alpha ^{\ast } \right ) \right ] \,\left ( \theta -\theta _ { 0}\right ) \\ & & + o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| ^{2}\right ) \\ & & \\ & = & \!\!\frac{1}{\sqrt{n}}\ \left ( \theta -\theta _ { 0}\right ) ^{t}\,v_{n}+\left ( \theta -\theta _ { 0}\right ) ^{t}\left [ \partial _ { \theta } \widetilde{s}_{1}\left ( \theta _ { 0},\widetilde{\alpha } _ { n}\right ) -\partial _ { \theta } \widetilde{s}_{1}\left ( \theta _ { 0},\alpha ^{\ast } \right ) \right ] \!-\left ( \theta -\theta _ { 0}\right ) ^{t}\,w_{n}\,\left ( \theta -\theta _ { 0}\right ) \\ & & + o_{p}\left ( \left\| \theta -\theta _ { 0}\right\| ^{2}\right ) , \end{aligned}\]]uniformly in @xmath15 in @xmath217 neighborhoods of @xmath554 , where @xmath555(here , @xmath556 and @xmath557 are vectors in @xmath558 while @xmath559 and @xmath560 are @xmath446 matrices ) . next , write @xmath561\]](@xmath562 denote the partial derivative with respect to @xmath563 with @xmath564 between @xmath78 and @xmath565 , and notice that by the definition of @xmath79 as the logarithm of a lefn density , for any @xmath42 @xmath566 = e\left [ \partial _ { \theta } \partial _ { \alpha } \psi \left ( y , r\left ( z^{t}\theta _ { 0};\theta _ { 0}\right ) ; \alpha \right ) i_{a}\left ( z\right ) \right ] = 0.\]]consequently , @xmath567 uniformly in @xmath442 deduce that for all @xmath568 in probability , @xmath569 use this fact and the order of @xmath570 to write @xmath571uniformly in @xmath524 , @xmath525 and @xmath78 in @xmath217 neighborhoods of @xmath572 by little algebra , @xmath573\]]@xmath574 .\]]deduce from the assumptions that @xmath575 converges in distribution to @xmath576 and @xmath577 in probability . finally , deduce that @xmath160 has the same asymptotic distributions as the maximizer of the quadratic form ( [ quad1 ] ) .
more precisely , apply theorems 1 and 2 of sherman ( 1994a ) to obtain first , the @xmath286consistency of @xmath160 and next , the asymptotic normality @xmath578 _ part ii : the behavior of _ @xmath253 _ _ _ _ by taylor expansion we can write @xmath579 where @xmath580 is independent of @xmath76 , @xmath581 \ i_{a}\left ( z_{i}\right ) , \ ] ] @xmath582 ^{2}\ i_{a}\left ( z_{i}\right ) .\ ] ] by lemma [ mise ] , @xmath583 , uniformly over @xmath584 with @xmath585 @xmath317 defined in ( a1a2 ) .
moreover , by lemma [ mise2 ] , @xmath586 uniformly over @xmath587 finally , recall that @xmath588 + \left [ r_{1}\left ( \theta _ { 0},h;\widetilde{\alpha } _ { n}\right ) -r_{1}\left ( \theta _ { 0},h;\alpha ^{\ast } \right ) \right ] .\ ] ] the order of @xmath589 and @xmath590 can be obtained in the same way as the order of @xmath591 taking the differences @xmath592 vanishes the constants of the dominating terms containing @xmath76 .
thus , up to terms independent of @xmath76 , the second bracket is negligible compared with @xmath593 , uniformly in @xmath545 on the other hand , by lemma [ rr ] , the first bracket is of order @xmath594 , uniformly in @xmath15 in @xmath595 neighborhoods of @xmath7 and @xmath524 and uniformly in @xmath62 in @xmath596 neighborhood of @xmath597 since @xmath160 was shown to be @xmath286consistent , deduce that @xmath174 is asymptotically equivalent to the maximizer of @xmath598 more precisely , @xmath259 in probability , where @xmath599 to close the proof it remains to show that @xmath227 defined in ( [ deeff ] ) is asymptotically equivalent to the maximizer of the objective function @xmath600 in equation ( [ infeasi ] ) .
for this we use inequality ( [ ikj ] ) and lemma [ idic ] .
moreover , we can decompose @xmath601 in the same way as @xmath602 and obtain the same orders , uniformly over @xmath603 uniformly with respect to @xmath62 and uniformly in @xmath604 $ ] , for some small @xmath605 .
note that @xmath606 shrinks to the set @xmath607 as @xmath608 therefore , the constants appearing in the dominating terms of the decomposition of @xmath609 vanishes as @xmath119 , provided that @xmath610 = 0 $ ] .
consequently , the @xmath611 orders are transformed in @xmath612 orders and thus @xmath613 uniformly in @xmath525 , @xmath524 and @xmath78 a sequence convergent to @xmath614 in probability , provided that @xmath119 .
the proof is complete .
newey , w.k .
efficient estimation of models with conditional moment restrictions , in g.s .
maddala , c.r .
rao and h.d .
vinod ( eds . ) _ handbook of statistics , vol .
11 , _ pp .
419- 454 , new - york : north - holland .
newey , w.k . and mcfadden , d. ( 1994 ) . large sample estimation and hypothesis testing , in r.f .
engle and d.l .
mcfadden ( eds . ) _ handbook of econometrics , vol .
iv , _ pp .
2111- 2245 , new - york : north - holland . | we propose a two - step pseudo - maximum likelihood procedure for semiparametric single - index regression models where the conditional variance is a known function of the regression and an additional parameter .
the poisson single - index regression with multiplicative unobserved heterogeneity is an example of such models .
our procedure is based on linear exponential densities with nuisance parameter .
the pseudo - likelihood criterion we use contains a nonparametric estimate of the index regression and therefore a rule for choosing the smoothing parameter is needed .
we propose an automatic and natural rule based on the joint maximization of the pseudo - likelihood with respect to the index parameter and the smoothing parameter .
we derive the asymptotic properties of the semiparametric estimator of the index parameter and the asymptotic behavior of our ` optimal ' smoothing parameter .
the finite sample performances of our methodology are analyzed using simulated and real data .
* keywords : * semiparametric pseudo - maximum likelihood , single - index model , linear exponential densities , bandwidth selection .
-1.5 cm -1.5 cm |
the 10.6gev @xmath0 collider cesr at cornell university in ithaca , new york , has served as an outstanding particle physics research facility for over 20 years .
the measured peak luminosity is @xmath1 .
its unique detector , constructed and operated by the cleo collaboration , has been able to provide a wealth of results from its focus on b meson studies and a variety of other topics such as charm and tau physics . since the year 2000
, the cleo iii detector ( @xcite ) has operated with a 4 layer si vertex detector , and a particle identification system based on _ ring imaging cherenkov _ ( rich ) detectors . due to the greatly increased luminosity during the cleo iii phase ,
a new readout and computing system was designed to cope with higher demands ( @xcite,@xcite ) . throughout the development process
, object oriented principles were applied consequently , resulting in a coherent and easily extensible set of software components that now cover all aspects of cleo computing .
part of this is the cleo iii database domain , implemented in an object database management system ( @xcite ) .
six individual online databases for alarms , program code , histograms , run setup data , run statistics data , and constants , comprise this system . all are implemented as c++ database servers , accessible to the distributed software components via corba communication interfaces ( @xcite,@xcite ) .
two online databases , for run statistics and constants , are also used in the cleo _ offline _ analysis framework .
this framework is used for all cleo physics analysis tasks , such as detector calibration , track reconstruction , monte carlo data generation , and user analysis .
it is based on a lightweight on - demand data delivery approach , as described elsewhere in these proceedings ( @xcite ) .
the cleo iii phase has ended in 2003 .
the storage ring is in the process of being converted to a charm factory ( _ cesr - c_,@xcite ) , adapted for center of mass energies between 3gev and 5gev .
it will serve as a new research tool for precision charm quark measurements , searches for new physics , and @xmath2 lepton studies .
it is expected to provide an array of quantum chromo dynamics ( qcd ) results that permit comparison to the most recent generation of lattice qcd calculations with unprecendented precision . since the original facility was not constructed for this energy domain , a set of superconducting wiggler magnets
is installed to ensure luminosities around @xmath3 .
several of these devices are already in operation ; the upgrade will be completed in early 2004 . in parallel ,
the cleo detector , now called cleo - c , is undergoing modifications to allow operation at those lower energies .
unfortunately , shortly after its deployment , the cleo iii vertex detector developed severe particle detection inefficiencies .
this effect reduced its usability considerably by causing limitations in secondary vertex identification . in preparation for cleo - c
, it was replaced with a 6 layer conventional vertex drift chamber .
the data structure of a cleo constants object is determined by the user , who writes a definition file in c - style syntax , e.g. : .... name : richchannel { uint32 channeladdress uint16 crate selectby uint16 threshold uint16 pedestal } .... the cleo build system reads this file and autogenerates a number of source code files for object database storage , database server management , corba interfacing , client data access , and java classes for graphical user interfaces .
this code is compiled and made available to all higher level software ( online and offline ) in libraries and java archives .
a single constants dataset , called a _
constants version _ , consists of one of more instances ( lines ) of data in the form of the described structure , such as the four integers in the above example . in cleo
, some versions have reached ca .
230000 lines .
every dataset receives a unique _ version number _ when it is written to the database . in addition
, a usage tag and run validity range can be associated with each constants version .
users usually do not have write access to the constants database . instead
, a user can create new constants objects programmatically , store them as ascii file , and then submit them to the central constants management system for installation .
after applying some consistency checks and safeguards , that system writes the new data to the database and updates validity information . after 2 years of operations ,
the cleo constants database had proven to be a reliable design that was extendable for new data types .
it became clear that separate copies of the full database were needed for online operation , track reconstruction processing , public analysis , and remote site monte carlo generation .
for that purpose , the protected online database was designated to be the _ master database _ ,
i.e. the only existing copy available for write access
. a number of copies acted as _ slave mirrors _ , readonly versions that were updated from the master as needed .
this scheme allowed minimization of the load on the master , a requirement for smooth and independent online detector operation .
despite the success in some areas , experience revealed that the constants database system was designed in a way that did not match the predominant usage patterns .
performance degraded with increased database size , causing concern for online operation .
hard disk and cpu contention on the server nodes could lead to wait times of minutes for a single request .
the system could not be expected to scale well for the rising number of clients .
this situation led to the decision to rewrite all constants database access code to ensure future scalability and efficiency .
a data object that is stored in an object database requires a unique identifier ( object identifier , oid ) for data management within the database and for retrieval .
such an object can have almost any shape , typically expressed in a c - style structure ( or a c++-style class ) .
a certain overhead for oid and storage is necessary and typically amounts to 2030 bytes , in addition to the space needed by the object itself . retrieving
an object also entails a certain computing overhead that is mostly independent of object size . from this fact
we conclude that it is desirable to keep the number of objects retrieved in a single database transaction small .
this notion was not part of the original constants database design : a cleo constants version , with a structure as described in the above example , could have many lines .
the original system architects had implemented this as a series of many oids , one per line .
the consequence was considerable overhead in storage space and cpu time needed for retrieval : a download request always asked for a complete constants version , i.e. many oids .
in addition , an inefficient search algorithm that often had to iterate through a large number of objects contributed to a steady decline in download performance as data volume rose over time .
another limitation became relevant when new versions for one of the existing data types were created in greater numbers than originally foreseen .
since every object must have a unique oid , the number of objects in any part of a database is limited . in this case
, there could not be more than @xmath4 versions .
a workaround was possible , but only at the expense of more complicated data distribution and longer search times . in the data organization domain ,
the original system violated a basic rule of database design : data normalization : the data contents of a constants version was stored not only together with its version number , but also with its usage information ( usage tag and run validity range ) .
in such a setup , one is severely limited in how data access can be managed .
it is , for example , not possible to associate more than one validity range - usage tag combination efficiently with an object , since a search with a certain tag and run number requires examination of many data objects .
since cleo required several such combinations , a cumbersome set of separate structures had to be invented to provide this functionality .
the maintenance of those structures was particularly arduous .
* one oid per constants version * no linear search on retrieval * copious oid space for a large number of versions * constants data stored independently of usage information * navigational elements that allow flexible usage management incorporated into the basic design according to these requirements , the new design consolidated all lines of one constants version into a single database object .
this approach reduces a database query to the retrieval of one object ; download pattern and storage structure are exactly matched .
the new organization now permits very fast data retrieval through a set of management structures that can be traversed quickly . a lookup with a usage tag and run number
is organized so that a validity list is searched in a tag dictionary , which then holds a direct reference to the desired data object .
the need for linear object searches was completely eliminated by using indexed version lists and hash maps for fast lookup via dynamic usage tags .
loading an object with a specified version number is as easy as opening a version list and dereferencing it with the version number as array index . with the number of database
objects greatly reduced , the available oid space is now considerably larger than needed for the lifetime of the cleo experiment . at the same time , the space requirements are less than half of the original needs .
a remarkable success story of object oriented client - server design was also the fact that the client interfaces did not have to be modified at all , since the redesign could remain completely confined to the database server implementation .
this allowed a smooth transition of all running build and production systems . for an illustration of the performance improvement that was achieved through the redesign , see figure [ fig : averagetimes ] . while the average load time of a single constants version ( @xmath50.2s ) , requested by regular physics reconstruction and analysis programs , were already an improvement from intermediate changes , the new system was able to reduce this time by more than a factor of five .
the distribution is much narrower , which translates into a fast , responsive , and reliable behaviour .
since a single analysis job can request over 100 different constants versions , the impact on real world operation is dramatic .
the generic trend in the high energy and nuclear physics computing community to purchase commercial software for their ever growing production systems has generated many varied experiences .
physicists do not write their own corba implementations , but they have realized that their database needs may be different from what industry requires , and that developing solutions on their own is possible and may turn out to be the better way . of concern for the entire community
is the fact that the widely used commercial implementation of an object database management system by objectivity inc .
( @xcite ) is being deprecated by almost all major collaborations for their large scale data storage needs .
a variety of reasons is cited : * the physics community is reluctant to adopt software without access to the source code , since its computing systems are usually highly non - standard implementations that require customization .
* an object database does not match traditional data access and usage patterns . * large scale data storage with an object database requires a highly sophisticated and finetuned setup and time consuming expert operation . * financial means in public research are limited .
although those issues are of little direct consequence for the project described here , because of the much smaller database sizes , a decline in overall acceptance will have an impact on product pricing , availability , and support .
similarly , the situation continues to be challenging with regard to platform support .
third party software packages , such as corba implementations , make inroads into the linux world at a slow pace .
but the linux on intel platform is in the process of displacing all other hardware / operating system combinations in the field .
this situation has placed limitations on what can be done with commercial software , and may prove a confining factor in future upgrades of the cleo computing systems .
9 clns report 94/1277 .
, presentation at computing in high energy and nuclear physics 2000 , padova , italy .
, presentation at computing in high energy and nuclear physics 2001 , beijing , china .
objectivity object database , by objectivity inc . , version 5.2 .
see http://www.objectivity.com .
corba ( common object request broker architecture ) is an industry standard for remote method invocation , see http://www.omg.org .
visibroker by inprise , version 5.1 .
references can be found on http://www.inprise.com/visibroker . c. jones , _ reconstruction and analysis on demand _ , presentation at this conference , psn thjt002 . for a summary ,
see d.g .
cassel , cornell university , lepp , _ cleo - c and cesr - c : a new frontier of weak and strong interactions _ , proceedings of the international conference on the structure and interactions of the photon and 14th international workshop on photon - photon collisions ( photon 2001 ) , 301 - 308 . | after the successful conclusion of the cleo iii phase , the cleo experiment at the cornell electron positron storage ring cesr is preparing for its transition to cleo - c .
this new program contains a wide array of physics studies at @xmath0 collisions at center of mass energies between 3gev and 5gev that will provide new insights into qcd .
because the existing silicon vertex detector needed to be replaced within a short time , a 6 layer vertex drift chamber has been installed in spring 2003 . at the same time , the existing ring imaging cherenkov detector , along with a conventional drift chamber , e.m .
calorimeter and muon chambers , will continue to be part of the experiment .
the cleo constants database system must provide efficient access to detector and analysis constants for a wide variety of purposes , such as hardware configuration , online operation and monitoring , calibration , and the complete cleo analysis framework .
since the original project , implemented through the objectivity object database management system , did not meet those requirements , the system was redesigned from the ground up , using the same technology .
it is currently being introduced into the production environment without significant interruptions . in this presentation
, we will outline the specifications of the constants database system , and then report on the process that led to the redesign .
performance comparisons and insights on relevant design aspects will be shown . |
security is an important issue of qkd .
one protocol for qkd is said to be secure if it can either generate the key secretly or stops the protocol when any eavesdroppers appear . in this section
, we proof that our proposed protocol is secure , even if the shared quantum channels are public .
first , we consider the security of the random key which is encoded in the outcomes of _ bob _
s measurement and the calculation result of `` _ _ alice _ _
s measurement''assuming that she does not apply any local operations on her particle .
where the symbol `` ' ' implies that _ alice _ does not really apply this measurement and get the outcome , it is just an imaginary measurement .
alice _ has performed a local operation on particle @xmath44 and a bell operator measurement , no eavesdroppers can gain any information , although _ alice _ announces publicly her bell measurement result on particles @xmath4 and @xmath5 .
the reason is that no one except _
_ knows exactly the entangled states of particles @xmath4 and @xmath0 once _ alice _ performs a local operation on her particle . there is only a probability of @xmath45 to guess the correct epr state in which particles @xmath4 and @xmath0 are in .
if eve eavesdrops @xmath46 key bits , she tends to succeed with a probability of @xmath47 which approaches @xmath48 and @xmath49 is sufficiently large .
further , if eve is so clever that she shares the quantum channels with _
alice _ and _ bob _ , or replaces the particles in transit to _
alice _ and _
bob _ by the particles prepared by her , she can succeed in eavesdropping with probability @xmath50 or @xmath45respectively .
this has been certified by us in[14 ] .
then _ bob _ can send some results to _ alice _ to detect the action of eve and thus , eve can not gain any information of the random key .
now turns to the problem of the certain key which is encoded in the local operations performed by _
alice _ on her particle .
it is available to _ bob _ by comparing the state of particles @xmath4 and @xmath5 gotten by _ bob _ with the result announced by _
thus the security of the certain key depends on whether _ bob _ can secretly infer the correct corresponding state of particles @xmath4 and @xmath5 using the entanglement swapping between two initial shared entangled pairs . in our previous work[14 ] , we showed that it is impossible for any eavesdroppers to know the bell operator measurement results by innerspring the transmission .
more detail , eve can not determine the outcomes of the sender and receiver s measurements in which the key is encoded , no matter what methods eve may choose .
there is a probability of @xmath50 to gain the correct information on the key when eve tries to share the quantum communication channels with _ alice _ and _
bob _ , and a success probability of @xmath45 by replacing the particles transported to legitimate users by the particles accessible to her .
this is a useful conclusion even for this different protocol .
that is to say , eve can eavesdrop no information of the certain key .
this protocol is secure , even the shared entangled states are public ( in another words , everyone know that the _ alice _ and _ bob _ shared states are @xmath51 ) . if _ alice _ and _ bob _ do not publish their s measured result , eavesdropper ( _ eve _ ) can not get the information from the known entangled channels .
for example , both the quantum channels are the same _ bell _ states @xmath36 and @xmath37 . by the entanglement swapping calculation , we know the state of these entangled particles are so the four bell - operator measurement outcomes of alice are equally likely , each occurring with probability @xmath45 . yet the eavesdropper can not get any information , although she owned many entangled pairs in state @xmath53 . compare eq.[@xmath5 ] with eq.[@xmath2 ] , we found if there is an eavesdropper , there is only @xmath50 probability that the two result are same .
_ bob _ can send some result random to _ alice _
, then _ alice _ can know there is eavesdropper with different result , they give up this key .
eve _ can not gain any information about the key .
if _ eve _ shared the entangled pairs @xmath59 with _
alice _ instead of _ bob _ , _ eve _ shared entangled pairs @xmath60 with _ bob _ , before the key distribution , both _
alice _ and _ bob _ do not know this .
then the process of key distribution become @xmath61 from eq.(@xmath32 ) and eq.(@xmath1 ) , we know that , the probability , which _ bell _ operator measurement of particle @xmath4 and @xmath5 is same as particle @xmath63 and @xmath64 , is only @xmath45 , when bob send some his measurement results to alice , alice can find if there is eavesdropper .
this work was partially supported by the cnsf ( grant no.90203018 ) , the knowledged innovation program ( kip ) of the chinese academy of sciences , the national fundamental research program of china with no.001gb309310 , k. c. wong education foundation , hongkong , and china postdoctoral science foundation . | a new protocol for quantum key distribution based on entanglement swapping is presented . in this protocol
, both certain key and random key can be generated without any loss of security .
it is this property differs our protocol from the previous ones@xcite .
the rate of generated key bits per particle is improved which can approach six bits ( 4 random bits and 2 certain bits ) per four particles . in cryptography ,
a message is uneavesdropped to any eavesdroppers . to achieve this goal ,
the message is combined with a key to produce a cryptogram . for a cipher to be secure
, it should be impossible to unlock the cryptogram without the key . in this sense , the security of the cipher depends on the security of the key , which is difficult to be assured by classical means .
quantum cryptography , also called quantum key distribution ( qkd ) , is defined as a procedure , in which two legitimate communicators can establish a sequence of key bits secretly and any unauthorized user can be detected . since the * bb84 * protocol[1],the first quantum key distribution scheme , was proposed ,
various quantum encryption schemes have been proposed such as * b92 * protocol[2 ] , the epr protocol[3 ] and other protocols[4 - 11 ] .
all these protocols for qkd produce random key bits , the security of the protocols is assured by this way . here
we propose a new qkd scheme in which not only random key bits , but also certain key bits can be generated without any loss of security . in this paper , we present a new protocol for quantum key distribution based on entanglement swapping . in our proposed protocol ,
_ alice _ only transports @xmath0 particles to _ bob _ , she can share @xmath1 bits , which include @xmath0 certain bits and @xmath2 random bits , with _
bob _ secretly .
this protocol brings some merits over the previous protocols , for example , the use efficiency can approach @xmath3 to illustrate our new protocol for qkd , we explain the process of entanglement swapping first .
it has been proposed by zukowski et al.[13 ] for two pairs of entangled particles with each pair in one of the bell states .
entanglement swapping works as follows .
consider two pairs of entangled particles @xmath4 , @xmath0 , @xmath5 and @xmath2 , prepared in bell states respectively @xmath6 and @xmath7 .
if a bell operator measurement is performed on particles @xmath0 and @xmath5 , then we have @xmath8 as is obvious , from above equation , the four possible results @xmath9,@xmath10,@xmath11 and @xmath12 occur with same probability and the outcome of each measurement is purely random .
suppose that the result @xmath13 is obtained , the state of the pair @xmath4 and @xmath2 after the measurement is consequently @xmath14 . which means particles @xmath4 and @xmath2 become entangled although they have never interacted .
recently , the entanglement swapping between two pairs of qubits has been used to realize schemes for quantum key distribution[9 ] . which do not need the legitimate users to choose between possible measurements to generate the key and
assure the security .
but only random key bits can be produced in these schemes .
our scheme , illustrated in fig .
1 , can be described as follows .
_ alice _ has a epr source .
she can apply a local operation @xmath15 , where @xmath16 on her particles .
_ alice _ and _ bob _ agree beforehand as following encoding @xmath17 where@xmath18 and @xmath19 are bell states .
@xmath20 they can perform series of operations as following , \1 .
_ alice _ creates epr pairs and sends some of her particles to _
bob_.
_ alice _ creates two epr pairs @xmath21 and @xmath22 , she sent particles @xmath0 and @xmath2 to _ bob_,and tells _
bob _ the state in which the particles are .
\2 .
_ alice _ applies a local operation on particle @xmath4 .
let the initial state of particles @xmath4 and @xmath0 be @xmath23 , and the state of particles @xmath5 and @xmath2 be @xmath24 .
_ alice _ performs a local operation @xmath25 on particle @xmath4 and thus , the state @xmath26 is turned into @xmath27 .
\3 .
_ alice _ makes a bell operator measurement on particles @xmath4 and @xmath5 .
we assume that the result of _ alice _
s measurement as @xmath28 , then she can infer the state of particles @xmath0 and @xmath2 as @xmath29 by the following equation , @xmath30 \4 .
_ alice _ calculates which state of particles @xmath4 and @xmath5 should be without local operation applied on particle @xmath4 when the epr state of particles @xmath0 and @xmath2 is @xmath29 .
@xmath31 from eq.(@xmath32 ) , _ alice _ get the state of particles @xmath4 and @xmath5 as @xmath33 , compared this result with expression(@xmath5 ) , she decodes the bits as @xmath34 .
\5 .
_ alice _ tells _ bob _ she has made a bell operator measurement on her particles @xmath4 and @xmath5 , but without mention the result of her measurement through classical channel .
\6 .
_ bob _ performs a bell operator measurement on particles @xmath0 and @xmath2 and infers the outcomes of _ alice _ s measurement . from the calculation of entanglement swapping , we know , after _ alice _ has performed a bell operator on particles @xmath4 and @xmath5 , _ bob _ s measurement result of his particles should be @xmath35 .
note that the result of _ alice _ s measurement inferred by _
bob _ is the corresponding state determined using eq.(@xmath32 ) . because _
bob _
does not know the exact state of entangled pair @xmath4 and @xmath0 after _
alice _
subjects her particle @xmath4 to a local operation .
what he can do is to use known information : entangled pairs @xmath36 and @xmath37 , to calculate the state particles @xmath4 and @xmath5 is in . in this case
, it is @xmath38 , which can be decoded as @xmath34 .
7._bob _ asks _ alice _ s corresponding outcome of her bell operator measurement on particles @xmath4 and @xmath5 by classical channel .
he knows the bell operator measurement of particles @xmath4 and @xmath5 is @xmath28 .
\8 .
_ bob _ compares the measurement result of particles @xmath4 and @xmath5 announced by _
alice _ with his calculation result , and he will find what the _ alice _ s local operation on particle @xmath4 is @xmath39 , _ bob _ gets the local operation is @xmath25 .
then he knows the certain bits is @xmath40 . obviously ,
both _ alice _ and _ bob _ know the state @xmath41 and the imaginary state of particles @xmath4 and @xmath5 as @xmath38 secretly . thus they establish the random key bits @xmath34 at the same time .
_ bob _ makes another es calculation as eq.(@xmath32 ) , he obtain the imaginary states of particles @xmath4 and @xmath5 as @xmath42 .
he decodes the random bits as @xmath43 .
they perform above procedures on each group of particles repeat .
\9 .
_ bob _ told _ alice _ some his qubits , _ alice _ will be sure if there is any eavesdropper .
then _ alice _ and _ bob _ securely share the certain key bit and the random key . |
the low mass x - ray binary ( lmxb ) 4u 1636 - 53 is one of approximately 10 neutron star x - ray binaries which have revealed high frequency pulsations during thermonuclear bursts , hereafter `` burst oscillations . ''
the @xmath9 hz burst oscillations in 4u 1636 - 53 were discovered with the _ rossi x - ray timing explorer _ ( rxte ) by zhang et al .
these oscillations are almost certainly caused by rotational modulation of a nonuniform x - ray emission pattern on the neutron star surface . near the onset of bursts
this pattern likely takes the form of a hot spot ( or possibly a pair of hot spots ) induced on the neutron star surface by the ignition of nuclear burning ( see strohmayer , zhang & swank 1997 ) . in the cooling part of the burst
lightcurve the modulation pattern is less certain , but it may be induced by @xmath10-modes excited by the thermonuclear flash ( see heyl 2002 ) , or perhaps `` extreme weather , '' in the form of hydrodynamic vortices induced by thermonuclear heating ( see spitkovsky , levin & ushomirsky 2001 ) . in particular , the large modulation amplitudes , high coherence and long term stability of the frequency are consistent with the rotation scenario ( see strohmayer et al . 1998a ; strohmayer & markwardt 1999 ; muno et al . 2000 and giles et al .
2002 ) .
there are now three known accreting millisecond pulsars ; sax j1808.4 - 3658 ( wijnands & van der klis 1998 ; chakrabarty & morgan 1998 ) , xte j1751 - 305 ( markwardt & swank 2002 ) , and xte j0929 - 314 ( remillard , swank & strohmayer 2002 ) .
the latter two having been discovered only very recently .
their spin periods are 401 hz , 435 hz and 185 hz , respectively .
several thermonuclear bursts have been observed from sax j1808.4 - 3658 with the bepposax wide field cameras ( in t zand et al .
1998 ) . in the brightest of the three bursts seen from sax j1808.4 - 3658 in t zand et al .
( 2001 ) found evidence for the presence of 401 hz burst oscillations , providing additional support to the idea that burst oscillation frequencies are set by the spin frequencies of neutron stars .
strohmayer et al .
( 1998b ) suggested that observations of burst oscillations in bursts at different binary orbital phases might allow a determination of neutron star orbital velocities by measuring the orbital doppler shifts in the burst oscillation frequency .
recently , giles et al .
( 2002 ) have studied the distribution of the highest observed , or `` asymptotic , '' oscillation frequencies in a large sample of bursts from 4u 1636 - 53 .
their sample spans a time baseline of more than four years .
they find a high degree of frequency stability over this timespan and attempt to constrain the neutron star orbital velocity using a portion of the observed asymptotic frequency distribution . over the last few years
a new regime of nuclear burning on neutron stars has been revealed with the observation of very long ( 3 - 5 hr ) thermonuclear bursts from six lmxbs ( see cornelisse et al .
2000 ; strohmayer & brown 2002 ; wijnands 2001 ; kuulkers et al .
2002 ; cornelisse et al .
these bursts likely result from unstable burning of the ashes ( as for example , carbon ) of hydrogen and/or helium burning ( see cumming & bildsten 2001 ; strohmayer & brown 2001 ; kuulkers 2001 ) .
of the six superburst sources so far only 4u 1636 - 53 has produced more than one event .
strohmayer ( see strohmayer & brown 2001 ) and wijnands ( 2001 ) independently identified the superburst from 4u 1636 - 53 which ocurred on february 22 , 2001 ; the former in pointed rxte observations , the latter in the rxte / asm data . using the asm data wijnands ( 2001 )
also identified a similar event from 4u 1636 - 53 which occurred 4.7 years earlier , thus providing the first constraint on the recurrence time for superbursts .
only two superbursts have been observed in detail with large area , high throughput instrumentation ; a superburst from 4u 1820 - 30 ( strohmayer & brown 2002 ) , and the february 22 , 2001 event from 4u 1636 - 53 . both were observed with the proportional counter array ( pca ) on board rxte .
here we report the discovery of coherent pulsations at @xmath11 hz during the february 22 , 2001 superburst from 4u 1636 - 53 .
the plan of this paper is as follows . in
1 we briefly describe the general properties of the superburst and the data available for high time resolution studies .
we then discuss our discovery of the @xmath11 hz pulsations and describe in detail the time evolution of the pulsation frequency and the expected behavior given the known binary orbital ephemeris for 4u 1636 - 53 . in
3 we describe our phase coherent timing study and show that orbital modulation of the frequency fits the data extremely well and that the pulsation is coherent .
we close in 4 with a summary and discussion of the implications of our findings .
a detailed study of the spectral evolution and energetics of this superburst will be presented in a future publication .
as part of an approved observing program for 4u 1636 - 53 , rxte observations were being conducted on february 22 , 2001 when a several hour long superburst was observed . during these observations data
were sampled in two high time resolution modes , both with a time resolution of 1/8192 seconds .
the first was an event mode ( e_125us_64m_0_1s ) with 64 energy channels and the second a `` burst catcher '' mode ( cb_125us_1m_0_249_h ) designed to capture high time resolution lightcurves of bursts across the entire pca bandpass . due to the duration and intensity of the superburst ,
the telemetry rate was very high and portions of rxte s onboard data recorder , which operates as a circular buffer , were filled before they could be downloaded .
unfortunately , this resulted in the loss of some high time resolution data during the superburst .
a similar circumstance ocurred when rxte observed a superburst from 4u 1820 - 30 ( see strohmayer & brown 2002 ) . in spite of these difficulties ,
two intervals of high time resolution data were obtained ; the first beginning just before the peak of the superburst , and the second in the decaying tail .
data covering the entire burst were obtained in the pca standard modes .
standard1 mode data provides 1/8 second resolution lightcurves across the entire pca bandpass , while standard2 mode provides 129 channel spectra every 16 seconds .
these modes use much less telemetry capacity than the high time resolution modes , and are written to a different virtual channel in the data recorder which was not overwritten . the 2 - 60 kev lightcurve of the superburst from standard1 data is shown in figure 1a with 1 second time resolution .
the rise in count rate beginning near 5500 seconds in the figure represents the source coming out of earth occultation . for several minutes after this the intensity level was about 4,800 counts / sec , significantly higher than the persistent source rate in the previous orbit
this difference is much too high to be accounted for by variations in the pca background rate and suggests that the burst may have already begun within the earth occultation gap .
the sharp increase in count rate which follows is real and can be seen with 1/8 second resolution in figure 1b .
this `` precursor '' began at 16:52:12 february 22 , 2001 ( utc ) .
the precursor has a timescale characteristic of normal ( 10 - 20 s duration ) type i bursts from 4u 1636 - 53 , except that it shows an interesting double peaked structure .
this behavior appears similar to that seen in the superburst from 4u 1820 - 30 , which also showed a normal type i burst precursor at its onset ( see figure 2 in strohmayer & brown 2002 ) .
further study of these interesting aspects of the superburst will be described in a future publication .
the first high time resolution data interval began at @xmath12 utc on february 22 , 2001 and lasted @xmath13 seconds .
this interval , which spans the peak of the superburst , is marked by the first set of vertical dashed lines in figure 1a .
we began our timing study by computing fft power spectra of this interval .
we used 1024 second intervals of the full 2 - 60 kev bandpass event mode data to compute two power spectra , each with a nyquist frequency of 4096 hz .
we then searched these power spectra for significant signals in the vicinity of 581 hz .
it was clear almost immediately that the power spectrum of the first 1024 s interval contained a significant signal at @xmath14 hz .
the power spectrum of this interval , rebinned in frequency space by a factor of 8 , is shown in figure 2a and reveals a highly significant peak near 582 hz .
figure 2b shows an expanded view of the region around 582 hz , revealing a pair of peaks separated by @xmath15 hz .
the highest peak in this power spectrum has a single trial chance probability of @xmath16 , so there is no doubt about the detection .
the implied frequency width of @xmath17 hz indicates a quality factor , @xmath18 , which without any additional timing analysis , is about a factor of 5 larger than that seen to date in burst oscillations from normal type i bursts ( see strohmayer & markwardt 1999 ; muno et al . 2000 ) .
we did not detect the signal in the remainder of the first high time resolution interval .
the upper limit to the rms pulsed amplitude was @xmath19 .
the 2nd high time resolution interval is marked by the 2nd pair of vertical dashed lines in figure 1a .
we searched this interval but detected no signal similar to that found in the first interval .
the upper limit to the rms pulsed amplitude in this interval was @xmath20 . in order to investigate the time dependence of the pulsation frequency and to determine whether it persists throughout the entire interval we computed a dynamic power spectrum using the @xmath21 power statistic
( see buccheri 1983 ; strohmayer & markwardt 1999 ) . however , in anticipation of the need for a precise timing analysis , we first barycentered the data using the standard rxte software ( faxbary ) and the jpl de200 solar system ephemeris ( standish et al . 1992 ) .
this analysis revealed that the two peaks in figure 2 result from two segments of detectable pulsations during each of which the pulsation frequency is observed to increase in a roughly linear fashion .
the @xmath22 s interval which contains these two pulse trains is marked by the vertical dotted lines in figure 1a .
figure 3 shows a contour map of the dynamic @xmath21 ( see below ) spectrum as well as the pca countrate as a function of time . to produce figure 3 we used 64 s intervals to compute @xmath21 and we started a new interval
every 16 s. this figure shows that the pulsations persist for _ hundreds _ of seconds , much longer than the @xmath23 s pulse trains observed in normal bursts .
moreover , the overall frequency drift spans only about @xmath24 hz , which is very small compared with the few hz drifts observed in normal burst oscillations ( strohmayer & markwardt 1999 ; muno et al .
the magnitude of this drift is similar to what could be produced solely by the doppler motion of the neutron star in its binary orbit .
a simple estimate , just using the observed magnitude of the frequency drift , would require @xmath25 , which is equivalent to a projected velocity of @xmath26 km s@xmath2 .
this number underestimates the velocity , since we have ignored the rate at which the frequency changes .
nevertheless , it suggests that we may be seeing a coherent pulsation modulated by orbital motion of the neutron star .
the 3.8 hr orbital period of v801 arae , the optical companion to the neutron star in 4u 1636 - 53 , is well known from optical observations .
recently , giles et al . (
2002 ) have used new optical observations to test and slightly modify the ephemeris determined from all historical optical observations by augusteijn et al .
their resulting ephemeris : hjd = 2446667.3179(33 ) @xmath27 , predicts the epoch of maximum optical brightness in units of heliocentric julian day .
although the formal statistical error in projecting this ephemeris forward to the time of the superburst is of order @xmath28 , there is an @xmath29 systematic fractional phase uncertainty which results from the fact that the optical lightcurve does not have any sharp fiducial features and also has often shown night to night variations ( see giles et al .
2002 for a discussion ) .
we used this ephemeris to predict the orbital phase of the @xmath22 s interval during the superburst which shows pulsations .
hereafter all phases will be given as fractional phases ( ie .
ranging from 0 to 1 ) .
the interval begins at mjd 51962.710160410 , which corresponds to a fractional phase of @xmath30 relative to optical maximum ( phase zero is optical maximum ) .
one addtional piece of information is required to predict the corresponding frequency evolution , the relationship between the optical and dynamical ephemerides . for systems like 4u 1636 - 53 , which show sinusoidal optical modulations and which have inclinations @xmath31 the standard interpretation has been that optical maximum corresponds to superior conjunction of the optical secondary , that is , when the secondary is furthest from the observer . figure 4 shows the predicted frequency behavior versus orbital phase of a coherent pulsation associated with the neutron star ( as , for example , its spin frequency ) given this relationship between the optical and dynamical phases .
the predicted phase range of the superburst pulsation interval is denoted with vertical solid lines .
the predicted frequency evolution is strikingly similar to that observed , moreover , even an uncertainty in the phase of @xmath32 would not change the qualitative conclusion that the frequency _ should _ be increasing because of the orbital motion . the original augusteijn ( 1998 )
ephemeris predicts an orbital phase of 0.26 for the start of the oscillation interval .
this is still in the range of orbital phases which would predict an increasing pulsation frequency .
these considerations strongly suggest that the observed frequency variations are consistent with the orbital motion of the neutron star . since frequency drift is a common feature of normal burst oscillations , it is a fair question to ask whether the frequency variations observed in this superburst which we claim to be orbitally modulated could be mimicked by intrinsic drifts unrelated to the orbit .
we argue that this possibility is unlikely for a number of reasons .
first , an intrinsic drift would need to replicate the apparent orbital modulation at the phase of the observed burst . as we show below
, the observed frequency variation is well fit with a concave upwards curvature which is rarely , if ever , seen in normal burst oscillations ( see strohmayer & markwardt 1999 ; muno et al .
a second consideration is that the frequency drift observed in the 4u 1636536 superburst is much smaller than that of all other normal burst oscillations .
examining bursts from many different x - ray binaries , muno et al .
( 2002 ) found 68 with oscillations , and measured the frequency drifts .
of those , about 30 had oscillations which began more than one second from the burst onset . however , _
none _ had drifts smaller than @xmath330.3 hz .
therefore we put the chance of observing a 0.04 hz drift at 1 in 30 , or @xmath3397% confidence . to derive
a more robust estimate would require more knowledge of intrinsic frequency drifts associated with superbursts , however , with only a single example so far , this is not currently feasible .
the frequency drift in the burst oscillation discussed here is consistent with a range of orbital phase , whose size gives a rough estimate of the likelihood of a chance agreement with orbital modulation .
this likelihood is related to the accuracy with which the phase can be predicted based on the optical ephemeris and the length of the pulse train . obviously , the more of the orbit covered by the pulse train , the greater our confidence would be in an orbital modulation interpretation .
based on the @xmath34 uncertainty in phase noted above and the @xmath35 s pulse train , we can estimate that only about a 5% to 10% range of orbital phase could produce a signature consistent with the observed drift .
below we fit orbital models and derive a range of phase spanning about 6% of the orbital phase which can produce an acceptable fit .
so , assuming all the observed frequency drift is intrinsic and not orbital there is roughly a 6% chance that it could be consistent with the expected orbital modulation . in our opinion , however , this should be considered an upper limit , since it seems quite unlikely that all the observed drift is intrinsic . in summary
, the observed frequency drift appears to be well consistent with that expected based on the known orbital parameters of v801 arae .
a chance agreement assuming all the observed drift were intrinsic to the superburst can be ruled out with confidences at the 9497% level .
in order to test the orbital modulation conclusion more quantitatively we performed coherent timing analyses using the complex @xmath36 statistic ( see strohmayer & markwardt 1998 for an example of the use of the @xmath21 statistic in a similar context ) , where ; @xmath37 the @xmath21 statistic is defined as @xmath38 , and the phase angle is given by @xmath39 . here
@xmath40 , where @xmath41 is the frequency as a function of time , and @xmath42 are the observed x - ray event times .
we used a circular orbit model to describe the time variation of the pulsation frequency and phase , _ vis _ ; @xmath43 here , @xmath44 , @xmath45 , @xmath46 and @xmath47 are the barycentric frequency , the projected neutron star velocity , the orbital phase at the start of the pulsation interval , and the orbital period , respectively .
we used this model to predict the pulsation phase at time @xmath48 and we varied the parameters in order to maximize @xmath21 .
we used the data from the @xmath49 s segment shown in figure 3 , but we excluded events in an @xmath50 s segment between the two pulse trains in which we did not detect significant pulsations .
we fixed the orbital frequency @xmath47 , since it is known with high precision and we can not hope to constrain it further with such a short data segment , and we let @xmath44 , @xmath45 , and @xmath46 vary .
we began by making a grid search over a range of parameter space which produced , `` by eye , '' reasonably plausible solutions .
we then used a general function extremizer , tnmin ( see nash 1984 ; implemented in idl by c. m. ) , to optimize the solutions .
we double checked the maximization using a simplex method ( amoeba ) implemented in fortran .
the two methods were in excellent agreement .
we find that this model can describe the frequency and phase evolution very well .
our best fits have @xmath51 which represents an enormous improvement compared to a model with no frequency evolution . as the data segment is short compared to the 3.8 hour orbital period it is possible to find statistically similar solutions with a range of orbital parameters .
indeed , there is a broad ridge in @xmath21 space on which the three parameters are strongly correlated .
the best model in the sense of having the largest @xmath52 has @xmath53 hz , @xmath54 km s@xmath2 and @xmath55 .
however , models with 90 km s@xmath2 @xmath56 175 km s@xmath2 are not strongly excluded .
a look at figure 4 confirms why this is so .
as long as the fitted orbital phase range is such that the curvature is concave up , then the other parameters , @xmath45 , and @xmath44 can more or less adjust to approximate the observed frequency evolution .
given the systematic uncertainty of order 0.05 in the optical phase and the additional uncertainty of the exact relationship between optical and dynamical phase it is not possible to prefer _ a priori _ a particular @xmath57 , however , the range above is consistent with what is expected based on the optical ephemeris as long as optical maximum ocurrs within @xmath32 of superior conjunction of the secondary .
these considerations show that the data can be well explained by a range of reasonable orbits , however , because of the shortness of the pulse train we can not place tighter constraints on the orbital parameters using these data alone .
nevertheless the observed pulsations are highly coherent as we are able to track @xmath58 consecutive cycles . in figure 5
we compare the standard fft power spectrum with the @xmath21 power spectrum using our best orbital parameters .
the pair of weak broad peaks near 581.91 hz are from the standard fft analysis and are effectively the same as the pair of peaks in figure 2 .
the narrow peaks near 582.1 hz come from the @xmath21 spectrum and result from the coherent addition of the two pulse trains .
the sidelobe peaks are caused by the need to bridge the gap between the two pulse trains .
figure 6 shows the best orbital frequency model plotted over the dynamic @xmath21 contours , and demonstrates visually that it tracks the frequency vs time contours very well .
the success of the coherent timing analysis provides strong evidence that we are seeing the orbital modulation of coherent pulsations during the superburst .
we can also investigate the problem in the context of phase evolution as a function of time .
for a given set of model parameters we can compute the phase , @xmath59 , of each x - ray event .
we can then break up the full set of events into a set of @xmath60 bins and compute the phase @xmath61 for each bin .
the phases are simply given by ; @xmath62 where the index @xmath63 runs over all events in bin @xmath64 .
we can then compute @xmath65 , where @xmath66 is the average phase angle computed from all @xmath60 bins .
we can then minimize @xmath67 to find a best fitting model . for a coherent signal the error @xmath68 in the phase angle
is given simply by @xmath69 , a result which we have confirmed with monte carlo simulations .
this procedure is essentially equivalent to the phase connected timing analysis of burst oscillations described by muno et al .
( 2000 ) . to further explore the phase coherence of our orbital solutions we broke up our event data into 24 segments and carried out the @xmath67 minimization described above .
we found that the best orbit model was entirely consistent with the model obtained from maximizing the @xmath21 statistic .
the minimum @xmath67 was 23.3 , which for 20 degrees of freedom ( 3 parameters and 1 additional degree of freedom from calculating @xmath66 ) gives a statistically acceptable solution .
the resulting phase residuals and errors are shown in figure 7 in units of milliperiods .
the rms level of @xmath70 milliperiods is indicated by the dashed horizontal lines .
this is equivalent to @xmath71 @xmath72s .
although the residuals are unstructured in the sense they do not show any large systematic deviations , there is some weak evidence for phase jitter ( or timing noise in the parlance of radio pulsar timing ) with a scale of about @xmath73 milliperiods on timescales of @xmath0hundreds of seconds , particularly during the first pulse train .
jitter of a similar magnitude is also suggested by dynamic power spectral and phase coherent studies of some burst oscillations ( see for example , wijnands , strohmayer & franco 2000 ; muno , chakrabarty & galloway 2002 ) . to find the @xmath5 confidence limits on the model parameters we found where @xmath74 , which is appropriate for a 3 parameter fit .
this gives a @xmath5 confidence range for each orbit parameter of ; @xmath75 , @xmath76 , and @xmath3 .
we also compared the circular orbital model with a strictly linear increase of the frequency .
the linear model achieves only a minimum @xmath77 , indicating that the circular orbit model performs significantly better . quantifying the significance of the difference in @xmath67 with the f - test shows that the additional parameter using the circular orbit model is significant at the @xmath78 level .
we note that the sense of the improvement is such that frequency models with curvature ( such that the frequency vs time curve is concave up ) fit much better than linear models .
we next used our best timing solution to phase fold the data and obtain an average pulse profile which we show in figure 8 .
the units are counts and we repeat 2 cycles for clarity .
the pulse profile is consistent with a sinusoid with a half amplitude of 1% of the mean number of counts .
having computed a coherent solution we also made a sensitive search for harmonics and subharmonics , but we did not detect any higher harmonics nor the @xmath79 hz subharmonic ( see strohmayer & markwardt 1999 ; miller 1999 ; strohmayer 2001 ) . the @xmath5 confidence upper limits to the signal amplitude at the 290 hz subharmonic and 1,160 hz 1st
harmonic are @xmath80 and @xmath7 of the mean count rate ( rms ) , respectively .
the observation of coherent pulsations during a superburst from 4u 1636 - 53 provides additional strong evidence of a rapidly rotating neutron star in 4u 1636 - 53 . for example , our observation of @xmath81 pulsation cycles at @xmath11 hz is roughly analogous to observing a standard @xmath82 s period radio or x - ray pulsar for @xmath83 days .
there would seem to be little doubt that such stability can only reflect the spin period of the neutron star . in the remainder of this section
we discuss some of the implications of our findings .
our coherent timing analysis places a constraint on the barycentric frequency of the coherent pulsation in 4u 1636 - 53 of @xmath84 hz . this range is higher than any of the asymptotic burst oscillation frequencies measured for 4u 1636
- 53 by giles et al .
our @xmath5 upper limit is about 0.4 hz above the highest asymptotic oscillation frequency ( see giles et al .
2002 ) , but the limits on the barycentric pulsar frequency are reasonably consistent with the _ total _ observed spread around this peak ( see figure 6 in giles et al .
the fact that the implied pulsar frequency is within @xmath01 hz of all the measured asymptotic frequencies provides further evidence that the frequencies observed during the more numerous , short duration ( @xmath85 s ) thermonuclear bursts are indeed set by the rotational frequency of the neutron star .
however , since the distribution of _ all _ asymptotic burst oscillation periods from 4u 1636 - 53 are not strongly correlated with orbital phase there arguably must be an additional source of variation in the asymptotic frequencies of burst oscillations above that caused by orbital motion of the neutron star .
there has been much recent theoretical work to provide an explanation of the few hz frequency drifts observed in burst oscillations .
for example , cumming et al .
( 2002 ) have reevaluated their earlier ( cumming & bildsten 2000 ) theoretical calculations of spin down in the neutron star surface layers produced by hydrostatic expansion and angular momentum conservation during thermonuclear bursts .
they conclude that such expansion alone is probably not capable of accounting for the largest observed frequency drifts .
more recently , spitkovsky , levin & ushomirsky ( 2002 ) have investigated the effects of burst heating in the atmosphere and show that this can drive strong zonal flows which can also produce frequency drifts .
they suggest that some combination of radial uplift and horizontal flows may explain the observed drift .
heyl ( 2002 ) has recently proposed that @xmath10-modes excited by the thermonuclear burst may be responsible for producing the observed modulation , especially in the decaying tails , and that the motions of these waves relative to the star can produce the observed frequency drifts .
our results provide further impetus for such work .
the discovery of coherent pulsations argues compellingly for the spin interpretation , which can be summarized as follows .
the observed modulations are caused by the rotation of some anisotropic emission pattern on the neutron star surface .
near burst onset the pattern is most likely a localized `` spot '' ( or perhaps antipodal spots ) . at late times in the burst tails it may be produced by an oscillation mode , perhaps the @xmath10-modes suggested by heyl ( 2002 ) , or the hydrodynamic vortices proposed by spitkovsky , levin & ushomirsky ( 2002 ) . however , the observed frequency drifts and variations in the asymptotic burst frequencies must be accomodated within a succesful model .
it seems most likely that the observed frequency drifts result from the fact that the emission pattern is not fixed in the rotating frame of the neutron star .
oscillation modes , as for example the @xmath10-modes , are not fixed in the rotating frame , and thermonuclear flame fronts can move across the surface of the star .
in addition , thermonuclear heating drives both radial and horizontal motions , both of which may contribute to the @xmath86 frequency drifts .
the coherent pulsations observed during the superburst have an average amplitude of @xmath87 of the mean 2 - 30 kev countrate , smaller than is typical for burst oscillations from normal bursts .
this raises the interesting question ; what causes the flux modulations during the superburst and how does it relate to that which operates during normal bursts ? if a weak coherent modulation were always present with some fixed amplitude , then if the source intensity increased as during the superburst
the modulation could become detectable .
this seems unlikely , however , since there are intervals during the burst at roughly similar intensities which do not show detectable pulsations .
more likely , there is some intrinsic change in the modulation amplitude caused by the superburst .
there is good evidence that these events result from the release of thermonuclear energy at great depth ( @xmath88 g @xmath89 , see strohmayer & brown 2002 ; cumming & bildsten 2001 ) .
if the energy release and/or transport is anisotropic then a flux asymmetry at the surface could be produced .
alternatively , the burst flux might excite an oscillation mode ( as , for example , suggested by heyl 2002 ) . in either case , if such a process produced a weak @xmath87 modulation which varied in strength by about a factor of two , then the intervals of strongest modulation would be detectable , but because the spectral power varies as the square of the modulation amplitude , a modest drop in the amplitude would render the modulation undetectable .
further insight into these issues could result from a phase resolved spectral study , which we will pursue in subsequent work .
our coherent timing analysis places a constraint on the projected neutron star velocity . as mentioned above ,
a range of orbital velocities , initial phases and pulsar frequencies are allowed by the data .
giles et al . ( 2002 ) used the observed distribution of asymptotic burst oscillation frequencies to constrain the neutron star velocity .
they obtained a @xmath90 upper limit on @xmath91 km s@xmath2 using a subset of 18 bursts which showed a tight clustering in asymptotic frequency .
although our inferred velocity range is outside their limit , the difference may reflect the exclusion of some bursts from the analysis by giles et al .
indeed , the overall spread in frequency of _ all _ the bursts from 4u 1636 - 53 is not strongly at odds with the higher neutron star velocity inferred from our timing analysis here ( see for example figure 6 in giles et al .
this does suggest , however , that it will likely be even more difficult to infer reliable neutron star orbital velocities using normal burst oscillation frequencies than indicated by the giles et al .
( 2002 ) study .
a measurement of the neutron star orbital velocity has implications for the masses of the binary components in 4u 1636 - 53 .
although the system inclination is not precisely known , the lack of both eclipses and dips suggests that it can not be much greater than @xmath92 ( see for example , frank et al .
figure 9 shows the range of allowable masses for the components , with the secondary mass along the abscissa and the neutron star mass along the ordinate .
velocity contours are shown for the limits calculated from our orbital timing solutions .
contours are drawn for two inclinations , @xmath93 ( dashed ) and @xmath94 ( solid ) .
the allowed masses are those enclosed by the velocity contours . in accreting binaries the mean density of the donor star
is constrained if the orbital period is known and the star loses mass via roche lobe overflow ( see for example , rappaport , joss & webbink 1982 ) .
the constraint can be expressed as ; @xmath95 where @xmath96 , @xmath97 and @xmath98 are the orbital period in hours , the mass of the donor , and the radius of the donor , respectively . if one assumes a particular mass - radius relation for the donor star , then the above constraint yields a mass estimate .
if one takes the canonical mass - radius relation for main sequence stars ( that is , @xmath99 ) , then for 4u 1636 - 53 ( @xmath100 ) this gives @xmath101 .
if instead one uses an empirical mass - radius relation for zero age main sequence ( zams ) stars then the mass estimate is reduced somewhat to 0.34 @xmath102 0.37 ( see patterson 1984 ; smale & mukai 1988 ) .
note , however , that the empirical mass - radius relation in the range @xmath103 is quite poorly defined ( see patterson 1984 ; baraffe et al .
1998 ) . as can be seen in figure 9
the mass estimates above are more or less consistent with our derived limits , but perhaps uncomfortably so , since they cluster near the lower velocity limit .
kolb , king & baraffe ( 2001 ) have recently investigated mass estimates in compact binaries .
they show that both mass loss and evolution of the donor prior to the onset of mass exchange can bias the mass estimate .
they argue that typically the real mass will be less than the estimate obtained from the roche lobe constraint and an assumed zams mass - radius relation .
these effects , if present , would tend to worsen the agreement with our lower range of masses in figure 9 .
one can ask what mass - radius relation would yield a mass estimate more comfortably within the constraints in figure 9 . for example , to have a mass of 0.5 @xmath104 the donor s mass - radius relation would need to be @xmath105 .
this does not represent a large discrepancy from expectations , particularly given the uncertainties at the low mass end .
the constraints shown in figure 9 are purely statistical in nature .
it seems likely , however , that a systematic uncertainty may also be present because the observed pulse train spanned only about @xmath106 of the orbit , and there is some indication of phase jitter from our timing analysis .
these considerations suggests that there could also be a source of systematic uncertainty in the derived orbital parameters .
because of this , and the sizable uncertainties associated with deriving a mass estimate , for example , the assumed mass - radius relation and the orbital inclination , we do not think that our present mass constraints are strongly at odds with theoretical expectations for the nature of the companion star .
however , if further observations continue to favor a high mass for the companion , then it could be an indication that some of the observed frequency drift is intrinsic to the superburst and not caused by orbital modulation .
augusteijn , t. , van der hooft , f. , de jong , j.a . ,
van kerkwijk , m.h . , &
van paradijs , j. , 1998 , a&a , 332 , 561 baraffe , i. , chabrier , g. , allard , f. & hauschildt , p. h. 1998 , a&a , 337 , 403 buccheri , r. et al .
1983 , a&a , 128 , 245 chakrabarty , d. & morgan , e. h. 1998 , nature , 394 , 346 cornelisse , r. , kuulkers , e. , int zand , j. j. m. , verbunt , f. & heise , j. , 2002 , a&a , 382 , 174 cornelisse , r. , heise , j. , kuulkers , e. , verbunt , f. , & int zand , j. j. m. 2000 , , 357 , l21 cumming , a. & bildsten , l. 2000 , apj , 544 , 453 cumming , a. et al . 2001 , apj , submitted , ( astro - ph/0108009 ) frank , j. , king , a. r. & lasota , j .-
1987 , a&a , 178 , 137 giles , a.b . , hill , k.m . ,
strohmayer , t. e. & cummings , n. 2002 , apj , 568 , 222 heyl , j. s. 2001 , astro - ph/0108450 in t zand , j. j. m. et al .
1998 , a&a , 331 , l25 in t zand , j. j. m. et al .
2001 , , 372 , 916 kolb , u. , king , a. r. & baraffe , i. 2001 , mnras , 321 , 544 kuulkers , e. 2001 , the astronomers telegram , # 68 , 68 , 1 kuulkers , e. et al .
2002 , a&a , 382 , 503 markwardt , c. b. & swank , j. h. 2002 , iauc , 7870 miller , m. c. 1999 , , 515 , l77 morgan , e. h. & chakrabarty , d. 1998 , nature , 394 , 346 muno , m. p. et al .
2002 , apj , submitted ( astro - ph/0204320 ) muno , m. p. , fox , d. w. , morgan , e. h. & bildsten , l. 2000 , apj , 542 , 1016 nash , s. g. 1984 , siam j. num .
, 21 , 770 patterson , j. 1984 , apjs , 54 , 443 remillard , r. a. , swank , j. h. & strohmayer , t. e. 2002 , iauc , 7893 smale , a. p. & mukai , k. 1988 , , 231 , 663 spitkovsky , a. , levin , y. & ushomirsky , g. 2001 , apj , 566 , 1018 standish , e. m. , newhall , x. x. , williams , j. g. & yeomans , d. k 1992 , in explanatory supplememt to the astronomical almanac , ed .
p. k. seidelmann ( mill valley : university science ) , 239 strohmayer , t. e. & brown ,
e. f. 2002 , apj , 566 , 1045 strohmayer , t. e. 2001 , advances sp .
28 , nos 2 - 3 , 511 strohmayer , t.e . , zhang , w. , & swank , j.h .
, 1997 , apj , 487 , l77 strohmayer , t. e. , zhang , w. , swank , j. h. , white , n. e. & lapidus , i. 1998a , apj , 498 , l135 strohmayer , t.e . , zhang , w. , swank , j.h . ,
& lapidus , i. , 1998b , apj , 503 , l147 strohmayer , t.e . , & markwardt , c.b . , 1999 ,
apj , 516 , l81 wijnands , r. 2001 , , 554 , l59 wijnands , r. , & van der klis , m. 1998 , nature , 394 , 344 zhang , w. w. , lapidus , i. , swank , j. h. , white , n. e. & titarchuk , l. 1997 , iauc , 6541 figure 1 : lightcurve of the superburst from 4u 1636 - 53 observed on february 22 , 2001 ( top ) .
the data are the 2 - 60 kev pca countrates from the standard1 datamode . in the top panel
the time resolution is 1 s. the dashed vertical lines denote the intervals with high time resolution event mode data .
the @xmath49 s interval in which pulsations are detected is shown by the vertical dotted lines .
an exploded view of the sharp rise near the start of the burst is also shown ( bottom ) . here
the time resolution is 1/8 s. note the double peaked profile which shows timescales typically seen in normal type i x - ray bursts ( @xmath23 s ) .
time is measured from 15:19:00 utc on feb .
22 , 2001 .
figure 2 : power spectrum ( top ) of the first 1024 second data interval starting at 17:00:44 utc .
the spectrum was computed using the full bandpass event mode data .
the nyquist frequency is 4096 hz .
the frequency resolution is 1/128 hz .
the significant peak near 581 hz is evident .
an expanded view of the region around 582 hz is also shown ( bottom ) .
the frequency resolution in this plot is 1/256 hz .
two peaks separated by @xmath107 hz are clearly resolved .
figure 3 : dynamic @xmath21 power spectrum of the 800 s interval containing significant pulsations .
the @xmath21 data were computed using 64 s data intervals and a new interval was started every 16 s. contours at 16 , 18 , 20 , 22 , 25 , 30 , 35 , 40 , 45 , 50 , 60 , 70 , 80 , and 90 are shown .
two pulse trains are clearly evident , separated by a gap of @xmath108 seconds
. the pca countrate ( 4 detectors ) is also shown ( see right axes ) .
time is measured from @xmath109 utc on feb .
22 , 2001 .
figure 4 : predicted orbital frequency evolution of a coherent pulsation centered on the neutron star in 4u 1636 - 53 and the orbital phase range of the 800 second pulsation interval during the superburst .
we used the optical ephemeris from giles et al .
( 2002 ) along with the single assumption that optical maximum corresponds to superior conjunction of the optical secondary .
the vertical lines denote the phase range of the 800 s pulsation interval .
the curve shows the frequency modulation which would be produced with @xmath110 km s@xmath2 for the neutron star .
this is not intended to represent a fit to the observed frequency evolution , but rather to simply show the qualitative nature of the predicted frequency evolution .
even with an offset of the phase relationship between optical max and superior conjunction of @xmath111 the ephemeris predicts that a pulsation frequency associated with the neutron star should be increasing .
figure 5 : comparison of the fft power spectrum which assumes no frequency or phase evolution with the @xmath21 power spectrum computed using the best orbital frequency evolution model .
the weak , broad peaks to the left of the figure are from the standard fft power spectrum , while the sharp peaks to the right result from the phase coherent @xmath21 power spectrum .
the sidelobe peaks in the coherent spectrum result from the gap between the two pulse trains .
figure 6 : dynamic @xmath21 contours as a function of frequency and time . also shown is the frequency evolution from the best orbit model .
the solid curve shows the model with @xmath53 hz , @xmath112 km s@xmath2 and @xmath113 and has a @xmath114 .
see the caption to figure 3 for the reference epoch .
figure 7 : pulse phase residuals ( observed - model ) for our best fitting orbit model .
the rms deviation is @xmath115 milliperiods and is denoted by the dashed horizontal lines .
there is some modest indication of an additional phase jitter in the first pulse train .
see the caption to figure 3 for the reference epoch .
figure 8 : average pulse profile in the full 2 - 60 kev band produced by folding the two observed pulse trains with the best fitting orbit model . the mean number of counts , 447340 , has been subtracted .
the profile is sinusoidal and has a half amplitude of @xmath87 of the mean number of counts .
phase zero is arbitrary .
figure 9 : allowable component masses for 4u 1636 - 53 inferred from the neutron star orbital velocity constraints .
a pair of velocity limits are plotted for two representative orbital inclinations , @xmath94 ( solid ) , and @xmath93 ( dashed ) .
the curves are labelled with their respective velocities . | we report the discovery with the proportional counter array on board the rossi x - ray timing explorer of highly coherent 582 hz pulsations during the february 22 , 2001 ( ut ) `` superburst '' from 4u 1636 - 53 .
the pulsations are detected during an @xmath0 800 s interval spanning the flux maximum of the burst . within this interval the barycentric oscillation frequency increases in a monotonic fashion from 581.89 to 581.93 hz .
the predicted orbital motion of the neutron star during this interval is consistent with such an increase as long as optical maximum corresponds roughly with superior conjunction of v801 arae , the optical companion to the neutron star in 4u 1636 - 53 .
we show that a range of circular orbits with velocity @xmath1 km s@xmath2 and fractional phase @xmath3 for the neutron star can provide an excellent description of the frequency and phase evolution .
the brevity of the observed pulse train with respect to the 3.8 hour orbital period unfortunately does not allow for more precise constraints .
the average pulse profile is sinusoidal and the time averaged pulsation amplitude , as inferred from the half amplitude of the sinusoid is @xmath4 , smaller than typical for burst oscillations observed in normal thermonuclear bursts .
we do not detect any higher harmonics nor the putative subharmonic near 290 hz .
the @xmath5 upper limits on signal amplitude at the subharmonic and first harmonic are @xmath6 and @xmath7 , respectively .
the highly coherent pulsation , with a @xmath8 provides compelling evidence for a rapidly rotating neutron star in 4u 1636 - 53 , and further supports the connection of burst oscillation frequencies with the spin frequencies of neutron stars .
our results provide further evidence that some millisecond pulsars are spun up via accretion in lmxbs .
we also discuss the implications of our orbital velocity constraint for the masses of the components of 4u 1636 - 53 . |
the trajectory of a particle embedded in a complex or even some seemingly simple structures may exhibit simultaneous modes of motion @xcite . an example is deterministic transport of a tracer particle in an infinite horizon ordered lorentz billiard , a set of fixed circular hard scatterers arranged in a square lattice @xcite .
a tracer moving freely among a set of scatterers and bouncing elastically once encountering one of them , exhibits intermittent behavior with long flights where the particle moves ballistically , separated by collision events which induce a diffusive like motion .
a similar behavior of the tracing particle can be induced by the flow acting in the phase space of chaotic hamiltonian systems @xcite .
the key to our discussion are power law distributed waiting times between collision events , induced by the geometry of the scatterers @xcite .
a condition for such non - drude like dynamics is that the radii of scatterers is smaller than half the lattice spacing and that the tracer is a point like particle , hence the tracer has an infinite horizon @xcite . also consider the very different type of motion of polymeric particles in living cells , where sub - diffusive motion is separated by long power law distributed flights which induce super - diffusion @xcite .
such systems are difficult to characterize since they exhibit at least two modes of motion .
a common tool in the analysis of such data is the spectrum of exponents @xmath6 @xcite .
one measures the @xmath1-th moment of the motion , for particles starting on a common origin @xmath7 with @xmath8 . here ,
for sake of simplicity we consider the one dimensional case and avoid other possible time dependencies like a logarithmic increase of moments with time .
scale invariant transport implies that @xmath9 is a constant independent of @xmath1 , for example , for brownian motion @xmath10 . in many fields of science one
finds that @xmath9 is a non - linear function of @xmath1 , the term strong anomalous diffusion is often used in this context @xcite .
surprisingly , in many cases the continuous spectrum @xmath6 exhibits a bi - linear scaling ( see details below ) .
examples for this piecewise linear behavior of @xmath6 include nonlinear dynamical systems @xcite , stochastic models with quenched and annealed disorder , in particular , the lvy walk @xcite and sand pile models @xcite .
recent experiments on the active transport of polymers in the cell @xcite , theoretical investigation of the momentum @xcite and the spatial @xcite spreading of cold atoms in optical lattices and flows in porous media @xcite further confirmed the generality of strong anomalous diffusion of the bi - linear type .
an example is a diffusive scaling @xmath11 below a certain value of @xmath12 and a ballistic scaling @xmath13 otherwise .
this is an example of a motion which is neither purely diffusive nor ballistic . while the mechanisms leading to strong anomalous diffusion could be vast ,
even the stochastic treatment of such systems is not well understood .
for example , diffusion equations , either normal or fractional @xcite , stochastic frameworks like fractional brownian motion or generalized langevin equations @xcite and the gauss - lvy central limit theorem @xcite fail to describe this phenomena .
we recently investigated the lvy walk process @xcite , a well known stochastic model of many natural behaviors which exhibits bi - linear strong anomalous diffusion .
examining four special cases we found that a non - normalizable density , with a diverging area underneath it , describes these processes @xcite .
this unusual infinite density , describes the ballistic aspects of strong anomalous diffusion .
here we find a general formula for the infinite density which , as we show below , is complementary to the gauss and lvy distributions .
the latter are the mathematical basis of many diffusive phenomena , and here we promote the idea that the infinite density concept is a rather general tool statistically characterizing the ballistic trait of the motion .
we show how these unnormalized distributions emerge from a basic model with wide applications , thus , possibly this overlooked measure may become an important tool . in mathematics the concept of a non - normalized infinite density
was thoroughly investigated @xcite while in physics this idea has gained interest only recently ( see below ) .
conservation of matter implies that the number of particles in the system is fixed , thus naturally one may normalize the probability density describing a packet of non - interacting diffusing particles @xmath14 to unity @xmath15 .
hence dynamical and equilibrium properties of most physical systems are described by densities which are normalized , for example , a boltzmann - gibbs state in thermal equilibrium , or solutions of boltzmann or fokker - planck equations with normalization conserving boundary conditions ( no absorbing boundaries ) . however , in some cases a closer distinction must be made , namely a distinction based on the observable of interest . probably the most common averaged observables are the moments @xmath16 of a process @xmath17 , so in this example the observable is @xmath18 and we will distinguish between high order moments @xmath3 and low order ones . in our case
we treat a system with mixed dynamics , and show that the low order moments are described by the standard machinery of non - equilibrium statistical physics ( fractional diffusion equations and lvy central limit theorem ) but the high order moments , which represent the ballistic elements of the process , are described by an infinite density . in this sense
the infinite density is complementary to the central limit theorem @xcite .
we find a general formula for this infinite density relating it to the velocity distribution of the particles , the anomalous diffusion exponent @xmath4 and the diffusion constant @xmath5 .
our work shows how the unnormalized state emerges from the norm conserving dynamics of the lvy walk .
previous work @xcite on applications of the infinite density in physics , dealt with bounded systems which attain an equilibrium .
for example the momentum distribution of cold atoms where the gibbs - measure is finite the infinite density describes the large rare fluctuations of the kinetic energy @xcite .
or intermittent maps with unstable fixed points , e.g. , the pomeau - manneville transformation on the unit interval @xcite . here
we consider systems that are not in equilibrium , and the dynamics is unbounded , showing that the potential applications of infinite densities are vast .
in the lvy walk process the trajectory of the particle @xmath17 consists of epochs of ballistic travel separated by a set of collisions which alter its velocity @xcite . at time
@xmath19 the particle is on the origin .
its initial velocity @xmath20 is random and drawn from a probability density function ( pdf ) @xmath21 , whose moments are all non - diverging @xcite .
the particle travels with a constant speed for a random duration @xmath22 whose pdf is @xmath23 .
the particle s displacement in this first sojourn time is @xmath24 . at time @xmath25 we draw a new velocity @xmath26 and a waiting time @xmath27 from the corresponding pdfs @xmath21 and @xmath23 .
the second displacement is @xmath28 .
this process is renewed .
the waiting times @xmath29 ( @xmath30 ) and the velocities @xmath31 ( @xmath32 ) are mutually independent identically distributed random variables .
the points on the time axis @xmath33 are the collision times , when the particle switches its velocity .
the position of the particle at time @xmath34 is @xmath35 here @xmath36 is the random number of collisions or renewals in the time interval @xmath37 .
the time interval @xmath38 is called the backward recurrence time @xcite . the last term in eq .
( [ eqmodel00 ] ) describes the motion between the last collision event and the measurement time @xmath34 @xmath39 a model where @xmath36 is fixed and @xmath34 is random , namely , a process stopped after @xmath36 collisions ( so @xmath40 ) is the classical problem of summation of independent identically distributed random variables for which the lvy - gauss central limit theorem applies @xcite . as already mentioned , we assume that @xmath41 hence from symmetry the density of particles @xmath14 is also symmetric since all particles start on the origin .
we also assume that the even moments of @xmath21 are finite , hence the tail of @xmath21 decays faster than any power law .
the pdf of waiting times is given in the long time limit by @xmath42 where @xmath43 and @xmath44 as discussed briefly in the summary , our main finding , that a non - normalized infinite density describes the density of particles @xmath14 , is found also for @xmath45 , but for simplicity of presentation we consider only a limited interval for @xmath4 .
the case @xmath46 corresponds to enhanced diffusion @xmath47 which is faster than normal diffusion but slower than ballistic @xcite .
the regime @xmath48 , is called the ballistic phase of the motion since then @xmath49 . for that case and for a general class of the velocity pdf , @xmath21 , the pdf @xmath14 is given by a formula in @xcite .
the case @xmath50 corresponds to normal diffusion in the mean square displacement sense @xmath51 .
a convenient tool is the laplace transform of the waiting times pdf denoted @xmath52 for a power law distributed waiting time under investigation , i.e. , @xmath46 , and for small @xmath53 the following expansion holds @xcite @xmath54 where @xmath55 is the averaged waiting time .
the position of the particle is rewritten as @xmath56 where @xmath57 are the flight lengths and @xmath58 . since the velocity distribution is narrow and symmetric , the pdf of flight lengths @xmath59 is also symmetric @xmath60 .
since the durations of the flights are power law distributed we have @xmath61 for large @xmath62 . in the regime
@xmath63 the variance of jump length is infinite hence the gaussian central limit theorem does not apply . in the regime
@xmath46 the average waiting time @xmath64 is finite . neglecting fluctuations , the number of flights is @xmath65 when @xmath34 is long .
hence in this over - simplified picture we are dealing with a problem of a sum of @xmath36 independent identically distributed random variables with a common pdf @xmath59 with a diverging variance . in this picture
the last jump @xmath66 is negligible .
thus , one may argue that @xmath67 is described by lvy s generalized central limit theorem .
this means that the pdf of @xmath67 is expected to be a symmetric lvy stable law @xcite ( see details below ) .
however , such a treatment ignores the correlations between flight durations and flight lengths and the fluctuations of @xmath36 .
in fact the finite speed of the particles implies that jumps much larger than the typical velocity times the measurement time @xmath34 are impossible .
thus , the mean square displacement and higher order moments always increase slower than ballistic , for example @xmath68 .
in contrast , a sum of independent identically distributed random variables with an infinite variance , often called a lvy flight , yields a diverging mean square displacement @xmath69 .
for that reason lvy walks , which take into consideration the finite velocity , are considered more physical if compared with lvy flights @xcite .
since both the gauss and the lvy central limit theorems break down in the description of moments of lvy walks , more specifically in the tails of the pdf of @xmath67 , it is natural to ask if there exists a mathematical replacement for these widely applied theorems .
a more serious treatment of the lvy walk is given in terms of the montroll - weiss equation .
let @xmath14 be the pdf of the position of particles at time @xmath34 for particles starting at the origin . from the mentioned symmetry of the process @xmath70
, @xmath71 so clearly all odd moments of @xmath14 are zero .
we define the fourier - laplace transform @xmath72 the montroll - weiss equation gives the relation between the distributions of the model parameters namely velocities and waiting times with @xmath73 @xcite @xmath74 here the averages are with respect to the velocity distribution @xmath75 . the derivation of this classical result is provided in an appendix .
note that in the original work of montroll and weiss a decoupled random walk was considered and the origin of eq .
( [ eqlw04 ] ) can be traced to the work of scher and lax @xcite and that of shlesinger , west and klafter @xcite ( see also @xcite ) .
examples @xcite for physical processes described by the lvy walk are certain non - linear dynamical systems @xcite , polymer dynamics @xcite , blinking quantum dots @xcite , cold atoms diffusing in optical lattices @xcite , intermittent search strategies @xcite , dynamics of perturbations in many body hamiltonian systems @xcite , and particle dynamics in plasmas @xcite . in what follows
we naively expect that one can reconstruct the normalized density of particles in the long time limit from the exact expressions for the moments of the process .
specifically , we obtain the exact expression for the long time limit of the integer moments of the lvy walk process @xmath76 with @xmath77 , where the case @xmath78 is trivial since @xmath79 , and then we use the moments to consruct a series equivalent to the fourier transform of the density @xmath80 luckily , we can evaluate analytically the sum for the model under consideration .
then we perform the inverse fourier transform of the such obtained function .
one then naively expects to get the long time limit of the density @xmath14 since we use the long time limit of the moments .
it turns out that this procedure yields a density which is _ not _ normalizable ( for reasons which will become clear later ) .
however , while the solution is not normalizable , it still describes the density of particles @xmath14 in ways which will hopefully become more transparent to the reader by the end of the manuscript .
to obtain the long time behavior of spatial moments @xmath81 we first find the laplace transforms @xmath82 . here , as mentioned , odd moments are zero due to the assumed symmetry @xmath41 . for that aim
we use the well known expansion @xmath83 p(x , u ) { \rm d } x= { 1 \over u } + \sum_{n=1 } ^\infty { ( i k ) ^{2 n } \langle x^{2 n } ( u ) \rangle \over ( 2 n ) ! } \ : .
\label{eqm01}\ ] ] expanding the numerator and denominator of the montroll - weiss equation , eq .
( [ eqlw04 ] ) using the expansion eq .
( [ eqmod02 ] ) , to the leading order in the small parameter @xmath84 , while keeping the ratio @xmath85 fixed , we get @xmath86}\ : \label{eqm02}\ ] ] with @xmath87 .
as is well known , such small @xmath53 expansions correspond to the long time limit @xcite . further expanding the denominator to find the first non - trivial term we obtain @xmath88 + \cdots \right\}\ : .
\label{eqm03}\ ] ] the leading @xmath89 term is obviously the normalization condition @xmath90 . in this expansion
we included terms of the order @xmath84 , while higher order terms , which are found from further expansion of the denominator in eq .
( [ eqm02 ] ) , but also non - universal terms which stem from the expansion of @xmath91 , eq .
( [ eqmod02 ] ) , to orders greater than @xmath92 , are neglected .
we taylor expand eq .
( [ eqm03 ] ) in @xmath85 , using the series expansion @xmath93 where @xmath94 is the pochhammer symbol .
averaging over velocities we find @xmath95 with @xmath96 .
comparing with eq .
( [ eqm01 ] ) yields in the small @xmath53 limit @xmath97 which is valid for @xmath46 and @xmath98 . using the laplace pair @xmath99 we find the long time limit of the even spatial moments @xmath100 with @xmath101 $ ] .
as well known , the process exhibits super - diffusion @xmath102 .
we now use the asymptotic result for @xmath103 to find the infinite density of the lvy walk process .
the density of particles @xmath14 and its fourier transform @xmath105 are defined according to @xmath106 we taylor expand @xmath105 as in eq .
( [ eqplan ] ) , yielding @xmath107 clearly , we have @xmath108 which is the normalization condition , namely , for any finite @xmath34 the density of particles @xmath14 is normalized to unity .
we next insert the exact long time expressions for the moments , eq .
( [ eq01 ] ) , in the series eq .
( [ eq04 ] ) , using @xmath109 to define @xmath110 here , the subscript @xmath111 denotes an asymptotic expression in the sense that we have used the long time limit of the spatial moments .
it is convenient to define @xmath112 hence @xmath113 it is easy to validate the following identity @xmath114 where @xmath115 with the expansion , @xmath116 , it is also easy to see that @xmath117 using mathematica this function is expressed in terms of the generalized hypergeometric function @xmath118\right\}/\alpha\ : .
\label{eq10a}\ ] ] for @xmath119 we find @xmath120 , thus the non - analytical behavior of @xmath91 for small argument @xmath53 , appears in @xmath121 when @xmath122 is large . for the inverse fourier transform we obtain @xmath123 for that aim we investigate @xmath124 using eq .
( [ eq10 ] ) we have @xmath125 the fourier pair of @xmath126 is @xmath127/2 $ ] , hence for @xmath128 @xmath129 note that this function is not integrable .
mathematically the integral formula of fourier transform eq .
( [ eq11 ] ) is valid for lebesgue integrable functions ( called @xmath130 ) while we are dealing with a distribution .
we insert eq .
( [ eq07 ] ) in eq .
( [ eq11 ] ) using the definition eq .
( [ eq08 ] ) , i.e. , @xmath131\right\}. \label{eq15}\ ] ] with the inverse fourier formula eq .
( [ eq13 ] ) the @xmath132-integration yields @xmath133 .
\label{eq16}\ ] ] the @xmath134 term is clearly arising from the normalization condition , namely the @xmath135 in eq .
( [ eq15 ] ) . hence using eq .
( [ eq14 ] ) , and the symmetries @xmath41 and @xmath136 we find @xmath137\ : , \label{eq17}\ ] ] for @xmath138 .
if we take @xmath34 to be large though finite , the function @xmath139 is not normalizable , since for small non - zero @xmath140 we get @xmath141 and thus the spatial integral over @xmath142 diverges .
since the number of particles is conserved in the underlying process , @xmath14 is normalized to unity .
one may therefore conclude that the non - normalized state @xmath142 does not describe the density of particles and hence does not describe physical reality .
this oversimplified point of view turns out to be wrong as we proceed to show in the next section .
we now define the infinite density . on right hand side of eq .
( [ eq17 ] ) , @xmath143 enters as a scaling variable , which we denote as @xmath144 clearly @xmath145 is the time average of the velocity and as usual we consider asymptotic long times .
the scaled variable @xmath145 has a non - trivial density , and it describes the ballistic scaling of the process @xmath146 . as we discuss below , there is not a unique scaling in the model , and the underlying process follows also a lvy scaling @xmath147 , which is a typical scaling for anomalous diffusion .
these two types of scalings are in turn related to the bi - scaling of the moments , found in many systems ( strong anomalous diffusion ) .
for now we define the infinite density @xmath148 in view of eq .
( [ eq17 ] ) we find our main formula @xmath149\ : , \label{eq20}\ ] ] where @xmath150 we will soon relate the infinite density @xmath151 , with the normalized lvy walk probability density @xmath14 while the definition eq .
( [ eq19 ] ) relates @xmath151 with @xmath142 .
* remark : * in our previous publication @xcite we called @xmath151 an infinite _ covariant _ density since the transformation of both space @xmath152 and time @xmath153 leaves @xmath154 unchanged . in mathematics
infinite invariant densities usually reflect solutions which are invariant under time shift only ( steady states ) , e.g. , for maps with an unstable fixed point . to avoid jargon we will call @xmath151 an infinite density meaning a non - normalized density . in the limit @xmath156 , eqs .
( [ eq20],[eq21 ] ) give @xmath157 this small @xmath145 behavior implies that the integral over @xmath158 diverges , hence the name infinite density .
one may rewrite eq .
( [ eqas01 ] ) in terms of the anomalous diffusion constant @xmath5 using @xmath159 with @xmath160 and @xmath161 .
the constant @xmath5 can be understood as a generalization of the standard diffusion constant in the framework of the fractional fokker - planck equation , @xmath162 @xcite ( note that this equation addresses the central lvy - like part of the pdf , @xmath163 , only , see sec .
`` the lvy scaling regime '' for a more detailed discussion ) .
briefly , it describes the width of the density field @xmath14 , so it is a measurable quantity .
further on , we find @xmath164 this relation provides the connection between the diffusive properties of the system and the infinite density . in principle
one may record in the laboratory the spreading of the particles , then estimate the exponent @xmath4 and @xmath5 by observing the center part of the packet , and afterward predict the small @xmath145 behavior of the infinite density . the relation given by eq .
( [ eqas04 ] ) is general in the sense that it does not depend on the particular form of @xmath21 . in the opposite limit of large @xmath145
we find @xmath165 where @xmath166 is the cumulative velocity distribution obeying , @xmath167 , as @xmath168 . to find this result we first integrate by parts eq .
( [ eq21 ] ) @xmath169 + \alpha \int_{\overline{v } } ^\infty v^{\alpha -1 } \left [ 1 - q(v)\right ] { \rm d } v \label{eqlv02}\ ] ] where we used @xmath170 , @xmath171 = 0 $ ] and @xmath172 . in the limit of large @xmath145
we may omit the second term on the right hand side of eq .
( [ eqlv02 ] ) if it is much smaller than the first . in that case
the following condition must hold @xmath173 { \rm d } v \over \overline{v}^\alpha [ 1 - q(\overline{v } ) ] } = 0\ : .
\label{eqlv3}\ ] ] with lhospital s rule we obtain the condition @xmath174 hence , if @xmath175 approaches zero faster than a power law for large @xmath145 the condition is met and @xmath176\ : .
\label{eqlv05}\ ] ] for example , it is easy to check this equation if @xmath177 for a certain @xmath178 and @xmath179 .
the equation is not valid for a power law behavior @xmath180 with @xmath181 , which is hardly surprising since we assume all along that moments of @xmath21 are either zero or finite , hence power law velocity distributions are ruled out from the start .
we then insert eq .
( [ eqlv05 ] ) in eq .
( [ eq20 ] ) to get eq .
( [ eqlv01 ] ) .
contrary to the small @xmath145 behavior of the infinite density , the large @xmath145 behavior is directly related to the velocity pdf , @xmath21 . in hindsight the asymptotic behavior eq .
( [ eqlv01 ] ) can be rationalized using a simple argument : the large _
@xmath182n_-th order moments @xmath76 are expected to depend only on the rare fluctuations , namely on the large @xmath145 limit of @xmath183 . with the definition of the infinite density we have @xmath184 .
we assume for now that we may replace @xmath142 with the density @xmath14 and get @xmath185 ( see reasoning for this below ) . in this case
the moments are @xmath186 for @xmath187 .
for @xmath78 the integral yields infinity as mentioned since @xmath151 is not normalizable . inserting the asymptotic approximation eq .
( [ eqlv01 ] ) we find by use of an integration by parts @xmath188 this is indeed the exact result eq .
( [ eq01 ] ) in the limit @xmath189 .
hence , the large @xmath145 behavior of the @xmath151 , eq . ( [ eqlv01 ] ) , yields the expected results for the high order moments of the spatial displacement .
we will soon justify the replacement @xmath190 with @xmath191 which led us to the correct result but first we turn to a few examples for the infinite density . _
( i ) _ for a two state model @xmath192/2 \label{eq22}\ ] ] we find using eqs .
( [ eq20],[eq21 ] ) @xmath193 & |\overline{v}| < v_0 \\ \ & \ \\ 0 & |\overline{v}| > v_0 .
\end{array } \right . \label{eq23}\ ] ] the particle can not travel faster than @xmath194 hence the infinite density is cutoff beyond @xmath194 and similarly @xmath195 beyond @xmath196 . _
( ii ) _ for an exponential velocity pdf @xmath197 @xmath198 \ : , \label{eq24}\ ] ] and @xmath199 is the incomplete gamma function . _
( iii ) _ for a gaussian model @xmath200 / \sqrt { 2 \pi } $ ] @xmath201\ : .
\label{eq25}\ ] ] _ ( iv ) _ while for a uniform model @xmath202 for @xmath203 and otherwise @xmath204 we obtain @xmath205 for @xmath206 .
similarly to the two state model , the infinite density @xmath151 for the uniform model is zero for @xmath207 , since the particle can not travel with a velocity greater than unity ( using the correct units ) . in fig .
[ fig1]-[fig3 ] infinite densities are plotted for several values of @xmath4 and for different models .
we now verify that the infinite density yields the moments @xmath103 with @xmath208 . using @xmath209 and eq .
( [ eq20 ] ) @xmath210 { \rm d } \overline{v } = \\ \ \\ 2 b t^ { 2 n + 1 - \alpha } \int_0 ^\infty \left [ \alpha { \cal f}_\alpha \left ( \overline{v}\right ) { { \rm d } \over { \rm d } \overline{v } } { \overline{v}^{2 n - \alpha } \over 2 n -\alpha } - \left ( \alpha - 1\right ) { \cal f}_{\alpha -1 } \left ( \overline{v } \right ) { { \rm d } \over { \rm d } \overline{v } } { \overline{v}^ { 2 n -\alpha + 1 } \over 2 n -\alpha + 1 } \right ] { \rm d } \overline{v}\ : .
\end{array } \label{eqrel02}\ ] ] integrating by parts and using eq .
( [ eq21 ] ) we find the moments , eq .
( [ eq01 ] ) .
of - course this is the expected result since we have constructed the infinite density with the long time behavior of the even moments .
we can , in principle , calculate the moments also from the normalized density because by definition @xmath211 the moments in eq .
( [ eqrel01 ] ) and eq .
( [ eqrel03 ] ) are identical in the long time limit , indicating that the infinite density and the density @xmath14 are related . rewriting eq .
( [ eqrel01 ] ) with the change of variables @xmath212 @xmath213 we conclude from comparison to eq .
( [ eqrel03 ] ) that @xmath214 for @xmath138 .
thus , we can use the density @xmath14 , obtained from a finite time experiment or simulation , to estimate with it the infinite density . in the limit @xmath215 the normalized density multiplied by @xmath216 and plotted versus @xmath143 yields the infinite density .
hence , the infinite density is not only a mathematical construction with which we may attain moments , rather it contains also information on the positions of particles in space and hence presents physical reality in the sense that it is experimentally measurable .
we now demonstrate this important observation with finite time simulations . obtained from numerical simulations for the model with uniform velocity distribution , @xmath217
$ ] . in the long time
limit simulations converge to the infinite density eq .
( [ eq26 ] ) .
we used function @xmath218 for @xmath219 as the waiting time pdf .
the parameters are @xmath220 , @xmath221 , and @xmath222 .
two histograms for @xmath223 and @xmath224 were sampled over @xmath225 realizations.,scaledwidth=50.0% ] [ fig1 ] $ ] , eq .
( 55 ) in the main text . @xmath221 and @xmath222.,scaledwidth=50.0% ] in fig .
[ fig1 ] we plot @xmath191 obtained from finite time simulations of the lvy walk , versus @xmath143 . according to the theory , in the limit @xmath215 this plot will approach the infinite density , eq .
( [ eq20 ] ) . such a behavior
is indeed confirmed with the simulations in the figure .
this implies that with a finite time simulation or experiment , which measures the density of particles , we can estimate the infinite density . in the figure
we see that data collapse ( for finite time ) is performing better for large @xmath143 , and convergence for small values of @xmath143 is slow .
indeed , for @xmath226 the density is well described by the lvy central limit theorem , as we soon explain . in the right panel in fig .
[ fig3 ] we demonstrate numerically the relation between the small @xmath145 behavior of @xmath151 and the anomalous diffusion coefficient @xmath5 , eq .
( [ eqas03 ] ) . plotting @xmath227 versus @xmath145 we see that in the limit of small @xmath145 , various models collapse on one master curve .
so with the knowledge of @xmath5 and @xmath4 , which as explained in the next section can be determined from data , we can predict the small @xmath145 behavior of the infinite density . as mentioned before , the large @xmath145 behavior of the infinite density , eq . ( [ eqlv01 ] ) , is sensitive to the shape of the velocity cdf @xmath228 , hence for large @xmath145 the corresponding curves in the figure depart . this sensitive tool in real experiments
may unravel important statistical information on the process .
practically , though dealing with long time limit solutions , eq .
( [ eqrel04 ] ) implies one should analyze the results for the shortest valid measurement times possible to maximize the probability of the region of interest .
thus measurement time must be long , so that asymptotic limit is reached , but not too long such that we can sample the tails of the pdf ( similar to many other problems in physics , for example large deviation theory ) .
the estimation of this time scale depends not only on the details of the model , but also on the number of particles undergoing the super - diffusive process , and the detailed problem of estimation of infinite densities from finite amount of data , is left for future work .
in contrast , the center part of the density , discussed in the next section , is described by the generalized central limit theorem , thus it is universal , but does not yield information on @xmath21 .
thus , infinite densities might become important tools in unraveling the origin of anomalous diffusion , a topic which attracted considerable theoretical attention , e.g. , @xcite and ref . therein . in simulations we reached a high degree of accuracy with extensive simulations and resolved the probabilities of the events at the pdf tails that are six order of magnitudes smaller than of those contributing to the pdf maximum ,
see the right panel of fig . 1 . the pdfs have been sampled with @xmath225 realizations .
the corresponding simulations were performed on two gpu clusters ( each consisting of six @xmath229 cards ) and took @xmath230 hours .
the main reason that we choose such a large number of particles , is to illustrate the theory with maximal possible ( at the moment ) precision .
single particle experiments in a single cell @xcite are conducted with much smaller number of particles , and for that reason it would be interesting to simulate the process for smaller numbers of particles and shorter times , to see if how accurately one may estimate the infinite density . however , in other experiments , like laser cooling of cold atoms @xcite , the number of particles is vast so these limitations are absent . *
remark : * upon inspection of fig .
3 , we realize that the probabilities of the events contributing to the pdf tails are six order of magnitude smaller than those contributing to the plotted pdf maximum .
this six order of magnitude difference is related to the choice of scale in the figure , since the infinite density diverges on the origin .
for example in the left panel we cut off the divergence on the origin . in reality , experiments are performed for finite times . hence the infinite density is slowly approached , but never actually reached , in the vicinity of the origin ( see fig .
1 ) . put differently , we need to sample rare fluctuations , and here sophisticated sampling algorithms could be useful @xcite . .
the right panel shows universal behavior of the infinite density in the small @xmath145 limit according to eq .
( [ eqas04 ] ) .
when the infinite density is rescaled with the anomalous diffusion constant @xmath5 times the constant @xmath231 these non - normalized densities fall on a master curve .
hence once we record @xmath5 and @xmath4 we can predict the small @xmath145 behavior of @xmath151 . here
@xmath232 and @xmath233 and @xmath234 for the two state , uniform , exponential and gaussian models respectively .
notice the log - log scale on the right panel .
, scaledwidth=90.0% ]
the key step in our derivation of @xmath235 was an expansion of the montroll - weiss equation , eq .
( [ eqlw04 ] ) , in the small @xmath236 limit while the ratio of @xmath85 is kept fixed .
this led to a scaling of the type @xmath146 .
as we showed , such an expansion leads to a non - normalized density @xmath151 , as opposed to the normalization dictated by eq .
( [ eqlw04 ] ) . to complete the the large order moments scheme we now consider the expansion of the montroll - weiss equation assuming that that @xmath237 is fixed and @xmath53 is small , i.e. , we consider the long time limit and inspect the pdf around the origin .
such an assumption means that we are looking for a solution with a scaling @xmath238 .
this scaling describes the center part of the pdf @xmath14 _ only _ and does address the pdf s tails .
it yields a normalized solution but at a price of a divergent second and all higher moments .
thus , if we wish to calculate the second moment @xmath239 , we need to estimate the tails of the @xmath14 and use the infinite density , i.e. perform the fixed @xmath85 expansion .
in contrast , if we want to estimate the normalization , or the low order moments @xmath240 with @xmath241 , we need an estimation of the center part of the pdf , and hence and expansion with @xmath237 fixed .
such low order moments can not be estimated with the infinite density because @xmath242 for @xmath243 , due to the singular behavior of the infinite density at the origin . to illustrate the above consideration
, we take eq . ( [ eqm02 ] ) and expand it assuming that @xmath244 and @xmath132 is small , @xmath245\right\rangle}{u - a_\tau \left\langle ( - i k v)^{\alpha } \left[1-\alpha\frac{u}{i k v } \right]\right\rangle } \sim \frac{1}{u - a_\tau \left\langle ( - i k v)^{\alpha}\right\rangle}. \label{p_center_ku11}\ ] ] the subscript
@xmath246 " indicates we aim at the center part of @xmath14 .
the term @xmath247 can further be simplified by recalling the symmetry of @xmath21 , @xmath248 by inserting it into eq .
( [ p_center_ku11 ] ) we find @xmath249 where @xmath5 is the anomalous diffusion coefficient , proportional to the moment @xmath250 of the pdf @xmath21 , eq .
( [ eqas03 ] ) .
thus , here the dynamics is not sensitive to the full shape of the velocity distribution but only to a particular @xmath4-th moment .
the solution eq .
( [ eqcen02 ] ) is well known @xcite ; its inverse laplace transform is @xmath251 as is well known , the fourier transform of the symmetrical lvy density @xmath252 is @xmath253 , which serves as our working definition of this stable density .
hence by definition , the inverse fourier transform of eq .
( [ eqcen05 ] ) yields a symmetric lvy stable pdf @xcite @xmath254\ : .
\label{equation}\ ] ] the lvy density is normalized , its second moment diverges since for large @xmath67 the solution has a fat tail @xmath255 . when @xmath256 we approach the gaussian limit . for the model with uniform velocity distribution ,
@xmath217 $ ] , see fig .
[ fig1 ] , plotted together with lvy distribution and rescaled infinite density .
the parameters are the same as in fig .
[ fig1].,scaledwidth=50.0% ] one can not say that either of the two montroll - weiss expansions presented so far is the correct expansion , without specifying the observable of interest and the domain in @xmath67 where one wishes to estimate the density . thus , if we wish to calculate the second moment @xmath239 from an estimation of the density of particles , we need the infinite density ( i.e. the fixed @xmath85 expansion ) estimating the tails of the @xmath14 . in contrast , if we want to estimate the normalization , or the low order moments @xmath240 with @xmath241 , we need an estimation of the center part of the packet , eq .
( [ equation ] ) .
as shown in fig . [ fig4 ] , for long , though finite times the center / tail region of @xmath14 is well approximated by the lvy / infinite density respectively .
we define a crossover position @xmath257 below / above which the lvy / infinite densities are valid approximations .
the maximum of @xmath14 is clearly on the origin ( neglecting possible finite time delta peaks found , for example , for the two state velocity model , since these do not contribute in the @xmath258 limit to the moments of the process ) .
hence , any estimation for the density for @xmath128 and long times must be smaller than the value of the density on the origin . on the origin @xmath259 with @xmath260 .
using @xmath261 we define @xmath262 via the condition @xmath263 since the transition to the lvy regime is found for the small argument behavior of the infinite density we use eq .
( [ eqas04 ] ) , @xmath264 and after some simple algebra we find @xmath265^{{1 \over 1 + \alpha } } \left ( k_\alpha t\right)^{1/\alpha}\ : .
\label{seccom02}\ ] ] though only a rough estimation for the transition , the important point is to notice that @xmath266 .
when data for @xmath14 is presented versus the scaled variable @xmath143 ( the variable @xmath145 ) we have a crossover velocity @xmath267 which goes to zero as @xmath215 since @xmath268 .
thus , the infinite density becomes a better approximation of @xmath191 versus @xmath143 as time is increased , see fig .
[ fig1 ] . based on this picture
let us calculate the @xmath1-th moment with @xmath269 , being a real number .
we divide the spatial integration into two parts and find @xmath270 \over \left ( k_\alpha t\right)^{1/\alpha } } x^q { \rm d } x } _ { \text { \mbox{inner l\'evy region } } } + \underbrace { 2 \int _ { x_c(t ) } ^\infty { { \cal i } \left ( { x \over t } \right ) \over t^\alpha } x^q { \rm d } x}_{\text { outer tail , infinite density region}}\ : .
\label{seccom03}\ ] ] changing variables according to @xmath271 in the first integral on the right hand side , and to @xmath272 for the second integral we have @xmath273 in the long time limit , the lower limit of the second integral @xmath274 while the upper limit of the first integral is a constant . for @xmath269
the second integral is by far larger than the first , hence we may neglect the inner region @xmath275 the @xmath276 in the integrand `` _ cures _ '' the non - integrable infinite density , i.e. , the non - integrability arising from the small @xmath145 divergence of @xmath151 , in the sense that the integral is finite when @xmath269 . when @xmath277 in eq .
( [ seccom05 ] ) with a positive integer @xmath278 we retrive eqs .
( [ eq01],[eqrel03a ] ) . in the opposite limit @xmath241
we find the opposite trend , namely now the inner region of the density @xmath14 is important in the estimation of the moments @xmath16 .
to see this we first note that in the intermediate region of @xmath67 , the lvy and infinite densities are equivalent .
the intermediate region are values of @xmath67 where the density @xmath14 is well approximated by the power law tail of the lvy density , before the cutoff due the finite velocity , and after the small @xmath67 region where the lvy density did not yet settle into a power law behavior .
indeed , there is a relation between the infinite density and the lvy density since they must match in the intermediate region .
using the known large @xmath122 behavior @xmath279 we obtain @xmath280 this is exactly the same as the small @xmath272 behavior of the infinite density , eq .
( [ eqas04 ] ) , hence , the two solutions match as they should . to obtain @xmath16 with @xmath241 we define a velocity crossover @xmath281 which is of the order of the typical velocity of the problem ( a velocity scale of @xmath21 ) .
for @xmath282 the density @xmath14 is well approximated by a lvy density ( since in the intermediate region the latter and the infinite density match ) , while beyond @xmath281 we have the infinite density description @xmath283 \over \left ( k_\alpha t \right)^{1/\alpha } } x^q { \rm d } x + 2 \int_{v_c t } ^\infty { { \cal i } \left ( { x \over t } \right ) \over t^\alpha } x^q { \rm d } x\ : .
\label{seccom06}\ ] ] now when @xmath215 and @xmath241 the contribution from the second integral is negligible and the upper limit of the first integral is taken to infinity , hence , after a change of variables @xmath284 , and using the symmetry @xmath285 @xmath286 we see that moments integrable with respect to the infinite density , i.e. @xmath269 are obtained from the non - normalizable measure . while for an observable which is non - integrable with respect to the infinite density , i.e. @xmath287 and @xmath241 , the lvy density yields the average and is used in the calculation .
further mathematical analysis of this behavior is required .
an observable like @xmath288 is non integrable with respect to the infinite density .
the average of @xmath289 with respect to the lvy density is finite .
hence the average @xmath290 is computed with an integration over the lvy density . on the other hand
consider an observable like @xmath291 where @xmath292 is the step function and @xmath293 .
now @xmath294 is integrable with respect to the infinite density but also integrable with respect to the lvy density .
so how do we obtain its average ?
do we use the infinite density or the lvy density ?
since @xmath14 is well approximated by the lvy density in the regions where the function @xmath294 is not zero , i.e. for @xmath295 the lvy density should be used in the calculation .
this example shows that not all averages of observables integrable with respect to the infinite density are obtained from the non - normalized density .
one can also construct an observable which is not integrable with respect to the lvy density neither with respect to the infinite density ; e.g. , a function that behaves like @xmath296 for large @xmath67 and @xmath297 for @xmath298 .
a trivial example is @xmath299 since the average of this function is given by both the lvy density to compute @xmath300 and @xmath301 which is found from the infinite density .
hopefully future rigorous work will present a full classification of observables and rules for their calculation .
another approach it to obtain a uniform approximation for @xmath14 based on the lvy and infinite densities .
this approximation which works for all @xmath67 will be presented elsewhere and it can serve as a tool for calculation of different classes of observables . from eqs .
( [ seccom05 ] , [ seccom07 ] ) we find @xmath302 with amplitudes @xmath303 and @xmath304 the amplitudes @xmath305 and @xmath306 diverge as @xmath307 from above or below , and in that sense the system exhibits a dynamical phase transition .
this behavior is shown in fig .
[ fig5 ] where finite time simulations show a clear peak of the moments amplitudes in the vicinity of @xmath308 .
the system exhibits strong anomalous diffusion with a bi - linear spectrum of exponents , i.e. , @xmath6 in eq .
( [ eqint01 ] ) is bi - linear .
such a behavior is demonstrated in fig .
[ fig6 ] , where @xmath6 versus @xmath1 is plotted using finite time simulations , which indicate that convergence to asymptotic results is within reach @xcite . as mentioned in the introduction ,
this piecewise linear behavior of @xmath6 , is a widely observed behavior .
, in the infinite long time limit these moments diverge as @xmath309 @xmath310 in this example ) .
lines correspond to eqs .
( [ secst02 ] , [ secst03 ] ) in the main text .
dots are the results of sampling with @xmath311 realizations . dashed line was obtained by combining together the lvy distribution and infinite density ( see fig .
[ fig4 ] ) and gluing them together at the point @xmath312 @xcite .
parameters are the same as in fig .
[ fig1 ] with @xmath313 .
, scaledwidth=50.0% ]
the lvy walk model exhibits enhanced diffusion where @xmath314 when @xmath46 . such a behavior is faster than diffusive and slower than ballistic . unlike gaussian processes with zero mean ,
the variance @xmath239 is not a sufficient characterization of the motion .
mono scaling solutions @xmath315 , which assume that the density of particles in the long time limit has a single characteristic scale , fail to describe the dual nature of the dynamics which contains both lvy motion and ballistic elements in it .
the density of particles at its center part is described by the symmetric lvy distribution .
if the lvy central limit theorem is literally taken , it predicts a divergence of the mean square displacement at finite times , i.e. , the variance of the stable pdf @xmath316 is infinite when @xmath63 .
the tails of the lvy walk model exhibit ballistic scaling , indeed as well known , the density is cutoff by ballistic flights . in this work
we have found the mathematical description of the ballistic elements of the transport .
an uncommon physical tool , a non - normalizable density describes the packet of particles .
the bi - linear scaling of the moments , i.e. , strong anomalous diffusion , means that we have two complementary scaling solutions for the problem .
the first describes the center part of the packet and is the well known lvy stable density with the scaling @xmath238 .
the second scaling solution describes ballistic scaling @xmath146 and is given by our rather general formula for the infinite density eq .
( [ eq20 ] ) .
the equation exhibits a certain universality in the sense that it does not depend on the full shape of the waiting times pdf .
it relates the non - normalized density with three measurable quantities : the velocity distribution of the particles , the anomalous diffusion exponent @xmath4 and the anomalous diffusion coefficient @xmath5 .
versus @xmath1 exhibits piecewise linear behavior .
the finite time simulations ( dots , parameters as in fig .
[ fig5 ] ) , for the uniform model , perfectly agree with theory ( lines ) : for @xmath241 @xmath317 otherwise @xmath318 ( see eq .
[ secst01 ] ) . here
@xmath319 . , scaledwidth=50.0% ] in this paper we have focused on a model with two scalings , namely ballistic @xmath146 and lvy @xmath238 motions . in real systems we may have a mixture of other modes of motion .
for example , gal and weihs @xcite measured the spectrum of exponents @xmath6 and found for large @xmath1 a linear behavior @xmath320 and @xmath321 so this motion is slower than ballistic but faster than diffusive .
this is probably related to the active transport of the measured polymer embedded in the live cell , namely to the input of energy into the cell . in this direction
, one should consider more general models .
for example variation of the lvy walks where waiting times are power law distributed but jump s lengths scale non - linearly with waiting times , i.e. @xmath322 with @xmath323 .
work on other regime of parameters is also in progress .
we recently investigated the regime @xmath50 , where the gaussian central limit theorem holds , and found that the infinite density concept remains in tact .
thus even a normal process , in the sense that the center part of @xmath14 is described by the gaussian central limit theorem , its rare fluctuations are related to a non - normalizable measure .
it is left for future rigorous work to see if infinite densities describe non - linear dynamical systems , at least those where power law distributions of waiting times between collision events , are known to describe the dynamics @xcite .
our results seem widely applicable . we know that bi - scaling of the spectrum of exponents is a common feature of different systems and strong anomalous diffusion is a well documented phenomena .
the lvy walk dynamics is ubiquitous and has been recorded in many systems .
hence we are convinced that the infinite density concept has a general validity , ranging from dynamics in the cell , motion of tracer particles in non - linear flows , spatial diffusion of cold atoms to name a few .
the question of estimation of infinite density from not too large ensembles of particles and finite time experiments is left for future work . in solid state physics , transport is in many cases characterized as either ballistic or diffusive .
the system falls into one of these categories depending on the ratio of the size of the system and mean free path . here
a completely different picture emerges .
depending on the observable , i.e. the order of the moment , the same system exhibits either diffusive or ballistic transport .
observables integrable with respect to the infinite density are ballistic , while observables integrable with respect to the lvy density exhibit super diffusion .
so in these systems the question of what is measured becomes crucial , and one can not say that the process / system itself falls into a unique category of motion .
at - least in our model , bi - scaling means that we have two sets of tools to master , the infinite density being the relatively newer concept which might need more clarifications in future work @xcite and become a valuable approach in other problems of statistical physics .
this work was supported by the israel science foundation ( ar and eb ) , the german excellence initiative `` nanosystems initiative munich''(sd and ph ) , and by the grant @xmath324 ( the agreement of august @xmath325 , @xmath326 between the ministry of education and science of the russian federation and lobachevsky state university of nizhni novgorod ) ( sd ) .
eb thanks the alexander von humboldt foundation for its support .
in this appendix we derive the montroll - weiss equation ( [ eqlw04 ] ) relating the fourier - laplace transform of @xmath14 with the velocity and waiting time pdfs , @xmath21 and @xmath23 respectively .
we denote the position of the particle @xmath327 ( instead of @xmath17 used in the text ) , where @xmath36 is the random number of collisions and rewrite eq .
( [ eqmodel00 ] ) @xmath328 the standard approach to the problem is an iterative approach , namely to relate the probability density of finding the particle at @xmath67 after @xmath329 collision events , with the density conditioned on @xmath330 collisions , e.g. @xcite . using the renewal assumption one
gets convolution integrals , and that leads to eq .
( [ eqlw04 ] ) .
we will use a slightly different method , to avoid a complete repeat of previous derivations , our approach is inspired by the renewal theory in @xcite .
the pdf of the position of the particles , all starting on the origin @xmath226 , at time @xmath34 is @xmath331 \rangle}\ : , \label{eq_pxt_vel}\ ] ] where @xmath134 is the dirac delta function . here
@xmath332 denotes the time of the @xmath36-th collision event @xmath333 .
the multiplication of the two step functions in eq .
( [ eq_pxt_vel ] ) , i.e. , the @xmath334 term , gives the condition @xmath335 , for the measurement time @xmath34 .
the summation over @xmath36 in eq .
( [ eq_pxt_vel ] ) is a sum over all possible number of collision events . transforming @xmath14 into the fourier - laplace domain , using eq .
( [ eqlw03 ] ) , we obtain @xmath336 the averages here are with respect to the velocity and the waiting time distributions . inserting eq .
( [ eqap01 ] ) in eq .
( [ eq_pku_vel1 ] ) we perform the time integral on the right hand side of eq .
( [ eq_pku_vel1 ] ) using @xmath337 and @xmath338 @xmath339}{u - i k v_n } { \prod_{j=1}^n \text{exp}[{(i k v_{j-1}-u ) \tau_j } ] } \right\rangle } \\ \\ & = \left\langle \frac{1-\text{exp}[{(i k v_0-u ) \tau_{1}}]}{u - i k v_0 } \right\rangle+\sum_{n=1}^{\infty } { \left\langle \frac{1-\text{exp}[{(i k v_n - u ) \tau_{n+1}}]}{u - i k v_n } \prod_{j=1}^{n } \text{exp } [ { ( i k v_{j-1}-u ) \tau_j } ] \right\rangle}\ : . \end{array } \label{eq_pku_vel2}\ ] ] here we separated the case of zero collisions @xmath340 from @xmath341 . because the velocities and waiting times are mutually independent , each separately being independent identically distributed random variables , we can perform the averaging . with the laplace
transform eq .
( [ eqmod01 ] ) , @xmath342 , we use @xmath343\rangle= \langle \hat{\psi } ( u - i k v ) \rangle^n\ : , \label{eqap03}\ ] ] where the remaining averaging on the right hand side is with respect to the velocity pdf only @xmath344 . hence from eq .
( [ eq_pku_vel2 ] ) we find @xmath345}{u - i k v } \right\rangle \sum_{n=0}^{\infty } { \left\langle \text{exp } [ { ( i k v - u ) \tau } ] \right\rangle^n}=\left\langle \frac{1-\hat{\psi}(u - i k v)}{u - i k v } \right\rangle \sum_{n=0}^{\infty } { \left\langle \hat{\psi}(u - i k v ) \right\rangle^n}\ : .
\label{eq_pku_vel3}\ ] ] d. n. armstead , b. r. hunt , and e. ott , _ phys .
e. _ * 67 * , 021110 ( 2003 ) .
d. p. sanders and h. larralde , _ phys .
e _ , * 73 * , 026205 ( 2006 ) . here
the authors demonstrate that even for a model of finite horizon , high order moments behave ballistically .
this is due to a non - trivial correlation between collision events , hence the infinite horizon condition is not a a necessary condition for our discussion .
m. courbage , m. edelman , s. m. saberi fathi , and g. m. zaslavsky , _ phys .
e. _ * 77 * , 036203 ( 2008 ) .
m. schmiedeberg , v. yu .
zaburdaev , and h. stark , _ j. stat
: theory and exp . _ ( 2009 ) p12020 ; deviations from biscaling are visible ( fig .
4 , therein ) , possibly originating from finite time effects .
p. siegle , i. goychuk , and p. hnggi , _ phys . rev
. lett . _ * 105 * , 100602 ( 2010 ) ; p. siegle , i. goychuk , and p. hnggi , _ epl _ * 93 * , 20002 ( 2011 ) .
linear generalized langevin equations with colored gaussian noise exhibit gaussian non - markovian anomalous diffusion ; it is still left to be shown if external non - linear fields may induce strong anomalous diffusion .
the lvy walk in its original schematization was equipped with a constant velocity of the motion , @xmath347 , i.e. , @xmath348 $ ] @xcite .
here we consider a generalization of the original model , so - called random walks with random velocities , where the velocity of the ballistic epochs is a random variable itself , see e. barkai and j. klafter , in _ chaos , kinetics and non- linear dynamics in fluids and plasmas _ , ed . by s. benkadda and g. m. zaslavsky , lect . not .
, v. 511 , pp .
373 - 393 ( springer berlin heidelberg , 1998 ) .
note that we use pdfs @xmath349 that are symmetric and their moments are finite .
thus we remain in the domain of levy walks , see also v. zaburdaev , m. schmiedeberg , and h. stark , _ phys . rev .
e _ * 78 * , 011119 ( 2008 ) .
n. korabel and e. barkai , _ phys .
_ * 108 * , 060604 ( 2012 )
. large deviation theory deals with remote tails of distribution functions , for example with the classical problem of summation of independent random variable with a common pdf . in this example
large deviations theory contains the gaussian central limit theorem and extends it to the regime of large fluctuations @xcite .
it is still left to be seen if large deviations theories can be used to obtain infinite densities for the lvy walk problem . | motion of particles in many systems exhibits a mixture between periods of random diffusive like events and ballistic like motion . in many cases , such systems exhibit strong anomalous diffusion , where low order moments @xmath0 with @xmath1 below a critical value @xmath2 exhibit diffusive scaling while for @xmath3 a ballistic scaling emerges .
the mixed dynamics constitutes a theoretical challenge since it does not fall into a unique category of motion , e.g. , the known diffusion equations and central limit theorems fail to describe both aspects . in this paper
we resolve this problem by resorting to the concept of infinite density . using the widely applicable lvy walk model , we find a general expression for the corresponding non - normalized density which is fully determined by the particles velocity distribution , the anomalous diffusion exponent @xmath4 and the diffusion coefficient @xmath5 .
we explain how infinite densities play a central role in the description of dynamics of a large class of physical processes and discuss how they can be evaluated from experimental or numerical data . |
the saturn nebula ngc7009 is one of the best - known planetary nebulae ( pne ) , and has been extensively studied both observationally and theoretically .
it is a large , double - ringed , high - surface - brightness pn , with a pair of low - ionization knots ansae along its major axis .
it has an h - rich o - type central star , with an effective temperature of 82,000k ( mndez , kudritzki & herrero @xcite ; kingsburgh & barlow @xcite ) .
ngc7009 has been the subject of many investigations since the early twentieth century .
berman @xcite made the first photometric measurements and isophotic contours .
spectrophotometric measurements extending to the ultraviolet were carried out in late 1930 s ( aller @xcite ) .
bowen & wyse @xcite and wyse @xcite obtained spectra of ngc7009 and estimated its chemical composition . in their work , nearly three hundred emission lines were detected , but only about 60 per cent were identified .
further studies were carried out by aller & menzel @xcite and aller @xcite .
ngc7009 is rich in emission lines , and is particularly well known for its unusually rich and prominent o ii optical recombination lines ever since the early high - resolution photographic spectroscopy observations in the 1930 s .
aller & kaler @xcite identified more than 100 o ii permitted transitions in the wavelength range 3100 4960 . in their longest - exposure spectrum of ngc7009 , lines as weak as 0.02 on the scale where @xmath12 = 100 were detected .
kaler & aller @xcite later reexamined the tracings of the long - exposure photographic plates used by aller & kaler @xcite and @xcite and reported several dozen additional very faint lines just marginally above the plate noise level .
barker @xcite obtained spectrophotometric observations at eight positions of ngc7009 , covering a wide wavelength range from 1400 to 10,000 .
he found that the c@xmath13/h@xmath14 abundance ratio derived from the c ii @xmath74267 optical recombination line ( orl ) is significantly higher than that derived from the c iii @xmath7@xmath71906,1909 collisionally excited uv intercombination lines , a phenomenon first discovered by perinotto & benvenuti @xcite . with the advent of modern high - quantum - efficiency and large - format linear detectors such as the ipcs and ccds , more and
more faint emission lines have been detected in the deep spectra of photoionized gaseous nebulae , including pne and h ii regions .
albeit faint , many of them are of important diagnostic value to probe various nebular atomic processes such as radiative and dielectronic recombination , continuum and bowen - like fluorescence and charge - exchange reactions .
a very detailed study of the o ii optical permitted lines in the deep spectra of ngc7009 was presented by liu et al .
@xcite who showed that for the 4f
3d transitions the departure from @xmath15-coupling is important .
they also found that the total elemental abundances of c , n , and o relative to hydrogen based on the recombination line measurements are about a factor of 5 higher than the corresponding values derived from collisionally excited lines ( cels ) , a discrepancy previously known to exist in the case of c@xmath13/h@xmath14 .
a number of postulations have been proposed to explain this discrepancy , including temperature fluctuations and density inhomogeneities ( peimbert @xcite ; rubin @xcite ; viegas & clegg @xcite ) , but failed to provide a consistent interpretation of all observations . a bi - abundance model given by liu et al .
@xcite , who postulate that nebulae contain h - deficient inclusions , provides a much better and natural explanation of the dichotomy . in this model ,
the optical recombination lines of heavy - element ions arise mainly from the cold " h - deficient component , while as the strong cels are emitted predominantly from the warmer ambient plasma of ` normal ' chemical composition .
deep spectroscopic surveys ( tsamis et al .
@xcite , @xcite ; liu et al .
@xcite , @xcite ; wang et al .
@xcite ) and recombination line analysis of individual nebulae ( liu et al . @xcite ; liu et al .
@xcite ; liu et al .
@xcite ; liu et al .
@xcite ) in the past decade has yielded strong evidence for the existence of such a cold " h - deficient component .
recent reviews on this topic are presented by liu ( @xcite and @xcite ) . in the study of emission line nebulae using faint heavy - element orls , which typically have intensities two to three magnitudes lower than h@xmath1 , reliable line identifications become an important issue .
correct identifications of spectral lines are fundamental to all spectroscopic studies . for lines commonly observed in astronomical spectra , a century of study has resulted in general agreement on transitions that give rise to strong lines observed at visible wavelengths . however , there is still much uncertainty about the proper identifications of many lines , particularly for fainter ones , and this problem is even more severe in other wavelength regions . as spectra approach fainter detection limits , the increasing number of features observed leads to a larger fraction of uncertain identifications .
the effort involved in assigning correct and astronomically sound line identifications for the large numbers of emission lines detected in high signal - to - noise ratio ( s / n ) spectra can be daunting .
recent notable work on this topic includes pquignot & baluteau @xcite , sharpee et al .
@xcite and zhang et al . @xcite .
in particular , sharpee et al .
@xcite developed a computer - aided code to identify lines detected in emission line objects .
the code automatically applies the same logic that is used in the traditional manual identification of spectral lines , working from a list of measured lines and a database of known transitions , and trying to find identifications based on the wavelengths and computed relative intensities of putative identifications , as well as on the presence of any other confirmed lines from the same multiplet or ion . while robust emission line identification is a difficult task , especially for faint lines , today deep , high - resolution spectra of pne ( liu et al .
@xcite , @xcite ; sharpee et al .
@xcite ) and h ii regions ( esteban et al .
@xcite , @xcite ; baldwin et al .
@xcite ) are routinely obtained .
since valuable information often results from the detection of previously unobserved low - abundant ionic species ( e.g. pquignot & baluteau @xcite ) , line identification remains a worthwhile investment .
emission line lists with robust identifications have been published for a number of bright pne and h ii regions , e.g. the orion nebula ( 3490 7470 ; baldwin et al .
@xcite ) , ic418 ( 3510 9840 ; sharpee et al .
@xcite , @xcite ) , ngc7027 ( 3310 9160 ; zhang et al .
@xcite ) . in the current work ,
we present deep spectra and identified emission lines of ngc7009 , another bright pn with archetypal rich and prominent heavy element orls .
we illustrate the techniques used to deblend faint lines , especially orls emitted by c@xmath14 , n@xmath14 and o@xmath14 ions .
a detailed analysis of those orls is the subject of a subsequent paper .
the spectra analyzed in the current work were observed from 1995 july to 2001 june , using the eso 1.52-m and the wht 4.2-m telescopes .
an observational journal is given in table [ observe_journal ] .
ngc7009 was observed in july of 1995 , 1996 , 1999 and june of 2001 , with the boller & chivens long - slit spectrograph mounted on the eso 1.52-m telescope .
all spectra were secured with a long slit , whose width could be varied as shown in table [ observe_journal ] . during the 1995 eso 1.52-m run ,
the b&c spectrograph was used with a ford 2048@xmath162048 15@xmath11m@xmath1615@xmath11 m ccd .
the slit was positioned at 79@xmath17 , i.e. along the nebular major axis and passing through the two outlying ansae ( c.f . the wfpc2 image of ngc7009 published by balick et al .
@xcite ) , and was offset 23 arcsec south of the central star ( cs ) in order to avoid the strong continuum emission from the cs , which , if included , would have had reduced the signal - to - noise ratios ( s / n s ) and the detectability of weak emission lines . the slit was about 3.5 arcmin long , and the slit width for all observations was 2 arcsec , except for one short exposure ( 60 s ) for which an 8 arcsec wide slit was used ( table [ observe_journal ] ) . during the 1996 run ,
the ccd on the b&c spectrograph was replaced by a thinned ultraviolet - enhanced loral 2048@xmath162048 15@xmath11m@xmath1615@xmath11 m chip of much improved quantum efficiency . for both observational runs , in order to reduce the read - out noise , the ccd was binned by a factor of 2 along the slit direction , yielding a spatial sampling of 1.63 arcsec per pixel projected on the sky .
several wavelength regions from the near ultraviolet ( uv ) atmospheric cut - off to approximately 5000 were observed using a 2400 line mm@xmath18 holographic grating , yielding a spectral resolution of approximately 1.5 fwhm .
lower resolution spectra from 3520 to 7420 were also obtained using a 600 groove mm@xmath18 grating .
the @xmath19 @xmath20 @xmath21 ( @xmath22 ) standard stars , feige 110 and ( the nucleus of the pn ) ngc 7293 , were observed with an 8 arcsec wide slit for the purpose of flux calibration . during the three nights observations in 1999 at the eso 1.52 m ,
three wavelength regions were covered : 8105 10,076 , 4697 6724 and 3965 4965 .
the spectral resolution for the three sets of spectra was about 3.0 fwhm .
the third night was cloudy and no standard star was observed , so the spectra of the range 3965 4965 were only wavelength calibrated .
they are of limited usage given the low s / n s .
data from the first ( 8105 10,076 ) and second ( 4697 6724 ) nights were better .
data reduction was no easy task , especially for data obtained in the first night , very accurate wavelength calibration was needed in order to satisfactorily subtract the many bright sky oh emission lines present in the spectra .
the first attempt to wavelength calibrate the spectra using arc lines of an hearfene lamp was not optimal as it was found that the geometric distortions of arc lines along the slit , which yielded crescent shaped line images with a curvature of about 1.2 pixels , were slightly different from those seen in the sky emission lines .
this probably resulted from the fact that the light path of the comparison lamp was not exactly the same as that of the sky light . as a result
, we opted to wavelength calibrate the spectra using the sky emission lines . the first attempt to wavelength calibrate a nebular spectrum using the sky emission lines detected in the same spectrum was not satisfactory , as for wavelengths longer than 9800 , no suitable sky emission lines not blended with nebular lines were available . eventually , we calibrated all the first night s nebular spectra using the sky emission lines detected in a single narrow slit exposure of the standard star feige 110 .
the second night s observations ( 4697 6724 ) were still wavelength calibrated using arc lines , as there were not enough number of sky lines in this wavelength region and the sky subtraction was less critical for this region .
despite all the efforts , subtraction of the many strong sky emission lines present in the 8105 10,076 wavelength region was still not entirely satisfactory .
observations at the wht 4.2-m telescope were obtained using the isis double spectrograph during two observing runs in 1996 and 1997 ( table [ observe_journal ] ) . for both the blue and red arms , a tek 1024@xmath161024 24@xmath11m@xmath1624@xmath11 m chip was used , yielding a spatial sampling of 0.3576 arcsec per pixel projected on the sky . in 1996 , gratings of 1200 and 600 groove mm@xmath18 were used in the blue and red arms , respectively .
the same set of gratings were used in 1997 .
two wavelength regions , 3618 4433 and 4190 4989 , were covered by the blue arm and another two wavelength regions , 5176 6708 and 6483 8005 , were covered by the red arm . during the 1996 run
, the slit was scanned across the whole nebula by uniformly driving the telescope differentially in right ascension ( ra ) .
the observations thus yielded average spectra for the whole nebula , which , when combined with the total h@xmath1 flux published in the literature and measured with a large entrance aperture , then yielded absolute fluxes of the whole nebula for all the emission lines detected in the spectra .
the spectra obtained in 1997 were secured with a fixed slit oriented at pa = 90@xmath17 and passing through the cs . for both wht runs ,
a 1 arcsec wide slit was used for nebular observations .
two @xmath22 spectrophotometric standard stars , bd+28@xmath17 4211 and hz44 , were observed using a 6 arcsec wide slit for the purpose of flux calibration .
there was still a gap in spectral coverage from 8000 to 8110 in the spectra described above .
this gap was filled with 2001 observations at the 1.52-m telescope , which covered a wide wavelength range from 7700 to 11,100 ( table [ observe_journal ] ) .
the spectral resolution was about 3.3 fwhm .
combined together , our observations of ngc7009 cover the complete wavelength range from the near - uv 3000 to the near - ir 11,100 .
all the spectra were reduced with standard procedures using midas .
the spectra were bias - subtracted , flat - fielded , cosmic rays removed and wavelength calibrated using exposures of comparison lamps , and then flux - calibrated using wide slit observations of spectrophotometric standard stars .
ozone absorption bands that affect data points shortwards of 3400 were corrected for using observations of the standard stars feige 110 and the cs of pn ngc 7293 taken with a 2 arcsec wide slit . as noted above , sky subtraction for spectra covering the wavelength range 8105 10,076 in 1999 with the eso 1.52-m telescope was not perfect .
the extracted one - dimensional ( 1-d ) spectrum of ngc7009 from 3040 to 11,000 is shown in fig.[spectra ] .
detailed spectral processing to deblend and identify weak lines is illustrated in the following sections .
analysis of important orls that are detected or deblended in the spectrum is the topic of a subsequent paper .
the logarithmic extinction at h@xmath1 , @xmath23(h@xmath1 ) , was derived by comparing the observed balmer line ratios , h@xmath24/h@xmath1 and h@xmath25/h@xmath1 , with the predicted case b values calculated by storey & hummer @xcite at @xmath8 = @xmath26 k and @xmath10 = @xmath26 cm@xmath3 .
this yielded a mean value of 0.174 , larger than 0.07 given by luo et al . @xcite but close to 0.2 by liu et al .
@xcite . as described in luo et al .
@xcite , the discrepancy is probably partly caused by the different regions of the nebula being sampled , although the possibility that it is caused by the calibration uncertainties can not be completely ruled out .
we have dereddened the observed line fluxes by @xmath27 where @xmath28 is the standard galactic extinction curve for a total - to - selective extinction ratio of @xmath29 = 3.1 ( howarth @xcite ) , and @xmath30 . [
cols="^,^,^,^,^,^,^,^ " , ] ( 1 ) : : blended with o ii
m89b 4f d[2]@xmath31
3d @xmath32d@xmath33 @xmath74669.27 , which contributes about 40 per cent to the blend at @xmath74669 .
fig.[histograms ] shows the histograms of intensity distributions of all the lines identified , as well as those of permitted lines from the most prominent heavy element ions , c ii , n ii , o ii and ne ii .
fig.[histo_cel_orl ] shows the distributions of the strengths of all identified orls and cels .
most orls as well as cels have intensities between 10@xmath4 and 10@xmath2 of h@xmath1 .
based on very deep , medium - resolution long - slit spectra of the saturn nebula ngc7009 , we have detected , deblended and estimated more than 1400 emission lines , with the weakest lines estimated down to intensity levels from 10@xmath5 to 10@xmath4 that of h@xmath1 . in total 1170 emission lines
are identified , both manually and systematically by the computer - aided code emili .
multi - gaussian line profile fitting is carried out across the whole wavelength range ( 3040 11,000 ) to retrieve intensities of blended features .
we first manually identify all the obvious emission features in the spectra , following the traditional method .
then we use multi - gaussian fitting to obtain fluxes of weak lines that are blended with strong features .
the most updated atomic transition data are utilized to estimate the fluxes of those weak lines whose reliable measurements are difficult . finally ,
all the aforementioned features , except those that are simply estimated from atomic data , are further identified with emili .
all the emission lines are presented in a comprehensive table with all detailed transition information included .
in addition to singly and doubly ionized species , recombination lines emitted by highly ionized ions of c , n and o are also identified .
we also detected forbidden lines from neutral species c i , n i and o i. the majority of lines are in the flux range 10@xmath4 10@xmath2 that of h@xmath1 . compared to several other pne that have been extensively studied in the recent literature , ngc7009 exhibits much more faint emission lines , thanks to its very rich and prominent orl spectra from abundant second - row elements .
the current set of spectra of ngc7009 are among the deepest ever taken for a pn .
the detection limit of the current data set is limited by the relatively low spectral resolution .
spectroscopy of this bright pn with better resolution will no doubt lead to direct detection of more and fainter lines .
cccllclllrrr + + @xmath34( ) & @xmath35 & @xmath36 & & i d & @xmath37( ) & mul . &
lower term & upper term & @xmath38 & @xmath39 & error ( % ) + + + + @xmath34( ) & @xmath35 & @xmath36 & & i d & @xmath37( ) & mul . &
lower term & upper term & @xmath38 & @xmath39 & error ( % ) + + 3041.49 & 0.2480 & 0.3170 & & si ii & 3042.19 & v14 & 3d 4d * & 4f 4d & 8 & 8 & : + & & & * & si ii & 3042.18 & v14 & 3d 4d * & 4f 4d & 4 & 4 & + 3043.75 & 1.3100 & 1.6737 & & ne iii & 3044.24 & & 3s 3d * & 3p 3d & 3 & 5 & 5.75 + & & & * & ne ii & 3045.56 & v8 & 3p 4p * & 3d 4d & 2 & 2 & + 3045.50 & 3.7120 & 4.7410 & & o iii & 3047.12 & v4 & 3s 3p * & 3p 3p & 5 & 5 & 3.03 + 3046.82 & 0.0950 & 0.1213 & & ne ii & 3047.56 & v8 & 3p 4p * & 3d 4d & 4 & 6 & 86.32 + 3058.27 & 1.0840 & 1.3814 & & o iii & 3059.30 & v4 & 3s 3p * & 3p 3p & 5 & 3 & 7.56 + & & & * & c v & 3057.50 & & 3s 3s & 3p 3p * & 3 & 5 & + & & & * & ne ii & 3057.83 & & 3d 2f & 5f 0@xmath403@xmath41 * & 6 & 6 & + & & & * & ne ii & 3057.87 & & 3d 2f & 5f 0@xmath403@xmath41 * & 6 & 8 & + 3060.68 & 0.1220 & 0.1554 & & [ cr iii ] & 3061.49 & & 3d4 1f & 3d4 1 g & 7 & 9 & 37.46 + 3107.84 & 0.1420 & 0.1794 & & [ ar iii ] & 3109.17 & & 3p4 3p & 3p4 1s & 3 & 1 & 35.21 + 3114.71 & 0.0870 & 0.1098 & & o iii & 3115.68 & v12 & 3p 3s & 3d 3p * & 3 & 1 & 34.48 + & & & * & fe ii & 3115.35 & v121 & c2 g & z2h * & 10 & 10 & + 3120.49 & 1.1510 & 1.4515 & & o iii & 3121.64 & v12 & 3p 3s & 3d 3p * & 3 & 3 & 4.34 + & & & * & fe i & 3121.15 & v163 & z7d & g5d & 3 & 5 & + 3124.33 & 0.0400 & 0.0504 & & ne ii & 3124.95 & & 3d 2f & 5f2@xmath403@xmath41 * & 6 & 6 & : + & & & * & ne ii & 3124.98 & & 3d 2f & 5f2@xmath403@xmath41 * & 8 & 8 & + 3131.70 & 29.9830 & 37.7420 & & o iii & 3132.79 & v12 & 3p 3s & 3d 3p * & 3 & 5 & 6.67 + 3137.60 & 0.0430 & 0.0541 & & ne ii & 3138.13 & & 3d 4f & 5f 2@xmath402@xmath41 * & 4 & 4 & 34.88 + 3148.23 & 0.0350 & 0.0439 & & ne ii & 3149.09 & & 3d 2f & 5f 2@xmath403@xmath41 * & 6 & 8 & 57.14 + & & & * & ne ii & 3149.04 & & 3d 2f & 5f 2@xmath403@xmath41 * & 6 & 6 & + 3161.95 & 0.0540 & 0.0677 & & c ii & 3162.66 & & 4f 2f * & 5d 2d & 6 & 6 & 27.78 + 3164.87 & 0.0800 & 0.1002 & & c ii & 3165.47 & v9 & 2p3 2d * & 3p 2p & 6 & 4 & 25.00 + & & & * & c ii & 3165.97 & v9 & 2p3 2d * & 3p 2p & 4 & 4 & + 3168.56 & 0.0580 & 0.0726 & & he i & 3169.02 & & 2s 1s & 15p 1p * & 1 & 3 & 25.86 + 3175.37 & 0.0290 & 0.0363 & & he i & 3176.27 & & 2s 1s & 14p 1p * & 1 & 3 & 24.14 + 3186.79 & 2.8410 & 3.5457 & & he i & 3187.74 & v3 & 2s 3s & 4p 3p * & 3 & 9 & 9.86 + 3194.37 & 0.0320 & 0.0399 & & ? & & & & & & & 31.25 + & & & * & ne ii & 3198.92 & v13 & 3p 4d * & 3d 4f & 4 & 4 & + 3202.24 & 4.8092 & 5.9883 & & he ii & 3203.17 & 3.5a & 3d 2d & 5f 2f * & 6 & 8 & 9.66 + 3209.95 & 0.0710 & 0.0883 & & [ fe ii ] & 3210.74 & & a6d & b2f & 4 & 8 & 21.13 + 3214.07 & 0.0670 & 0.0833 & & [ fe ii ] & 3214.67 & & ( 5d)4s 6d & ( 3d)4s 4d & 8 & 8 & 20.90 + 3217.25 & 0.1880 & 0.2336 & & ne ii & 3218.19 & v13 & 3p 4d * & 3d 4f & 8 & 10 & 9.95 + 3228.91 & 0.0580 & 0.0719 & & si ii & 3229.55 & & 3d 2p * & 4f 2d & 4 & 6 & 10.34 + 3229.87 & 0.0560 & 0.0694 & & n ii & 3230.54 & & 3d & 3p * & 7 & 5 & 10.71 + 3231.56 & 0.0500 & 0.0620 & & ? & & & & & & & 12.00 + 3243.22 & 0.0600 & 0.0743 & & ne ii & 3244.09 & v13 & 3p 4d * & 3d 4f & 6 & 8 & 10.00 + 3250.23 & 0.0210 & 0.0260 & & fe iii & 3250.84 & & 4 g 4d3i & 2i 5p3i * & 11 & 11 & 42.38 + 3257.97 & 0.0670 & 0.0828 & & [ mn ii ] & 3258.56 & & a7s & a3h & 7 & 11 & 15.82 + 3260.14 & 0.1490 & 0.1840 & & o iii & 3260.85 & v8 & 3p 3d & 3d 3f * & 5 & 7 & 8.86 + 3264.61 & 0.1160 & 0.1432 & & o iii & 3265.32 & v8 & 3p 3d & 3d 3f * & 7 & 9 & 9.48 + 3266.49 & 0.0170 & 0.0210 & & o iii & 3267.20 & v8 & 3p 3d & 3d 3f * & 3 & 5 & 52.35 + 3276.61 & 0.0120 & 0.0148 & & fe ii ] & 3277.35 & v1 & a4d & z6d * & 8 & 10 & 74.17 + 3286.78 & 0.0850 & 0.1046 & & [ fe ii ] & 3287.35 & & a6d & b4d & 2 & 6 & 12.35 + 3289.46 & 0.0220 & 0.0271 & & o ii & 3289.98 & v23 & 3p 4p * & 4s 4p & 2 & 4 & 40.91 + 3298.64 & 1.1070 & 1.3598 & & o iii & 3299.40 & v3 & 3s 3p * & 3p 3s & 1 & 3 & 9.79 + & & & * & ne ii & 3297.73 & v2 & 3s 4p & 3p 4d * & 6 & 6 & + 3303.62 & 0.0390 & 0.0479 & & fe iii & 3304.31 & & 2s 4s3s & 4d 4p3p * & 3 & 5 & : + 3311.61 & 3.0740 & 3.7695 & & o iii & 3312.32 & v3 & 3s 3p * & 3p 3s & 3 & 3 & 5.57 + & & & * & ne ii & 3311.27 & v2 & 3s 4p & 3p 4d * & 4 & 2 & + 3321.71 & 0.0480 & 0.0588 & & [ fe iii ] & 3322.53 & & 5d & 7s & 9 & 7 & : + 3323.74 & 0.0600 & 0.0735 & & cr ii & 3324.47 & & x6d * & 5d 9s6d & 8 & 10 & 66.67 + 3325.94 & 0.0460 & 0.0563 & & ne ii & 3327.15 & v2 & 3s 4p & 3p 4d * & 4 & 4 & 29.13 + 3328.47 & 0.0430 & 0.0526 & & ne ii & 3329.16 & v12 & 3p 4d * & 3d 4d & 8 & 8 & 31.40 + 3330.04 & 0.0120 & 0.0147 & & ne ii & 3330.73 & v20 & 3p 2d * & 3d 2f & 6 & 6 & : + 3331.31 & 0.0610 & 0.0746 & & [ fe ii ] & 3331.96 & & a6d & b2 g & 8 & 10 & 45.90 + 3334.16 & 0.3500 & 0.4279 & & ne ii & 3334.84 & v2 & 3s 4p & 3p 4d * & 6 & 8 & 8.46 + 3340.07 & 4.4230 & 5.4035 & & o iii & 3340.76 & v3 & 3s 3p * & 3p 3s & 5 & 3 & 6.50 + 3341.68 & 0.6390 & 0.7805 & & [ ne iii ] & 3342.50 & & 2p4 1d & 2p4 1s & 5 & 1 & 1.96 + & & & * & ne ii & 3344.40 & v2 & 3s 4p & 3p 4d * & 2 & 2 & + 3345.08 & 0.0110 & 0.0134 & &
ne ii & 3345.88 & v12 & 3p 4d * & 3d 4d & 6 & 4 & 36.36 + 3350.05 & 0.0360 & 0.0439 & & fe i & 3350.62 & & z5p * & 7/2@xmath403/2@xmath41 & 3 & 5 & 27.78 + 3352.47 & 0.0650 & 0.0793 & & [ cl iii ] & 3353.17 & & 3p3 4s * & 3p3 2p * & 4 & 2 & 15.38 + 3353.90 & 0.1403 & 0.1711 & & he i & 3354.55 & v8 & 2s 1s & 7p 1p * & 1 & 3 & 6.37 + 3354.52 & 0.1827 & 0.2228 & & ne ii & 3355.02 & v2 & 3s 4p & 3p 4d * & 4 & 6 & 6.02 + 3357.07 & 0.0170 & 0.0207 & & ne ii & 3357.82 & v12 & 3p 4d * & 3d 4d & 6 & 6 & 58.82 + 3359.95 & 0.0330 & 0.0402 & & ne ii & 3360.60 & v2 & 3s 4p & 3p 4d * & 1 & 3 & 30.30 + & & & * & o ii & 3360.62 & v52 & 3d 4f & 5p 4d * & 4 & 6 & + 3362.11 & 0.0220 & 0.0268 & & ne ii & 3362.94 & v12 & 3p 4d * & 3d 4d & 4 & 2 & 45.45 + & & & * & ne ii & 3362.71 & v42 & 3p 2p * & 4s 2p & 4 & 2 & + 3366.73 & 0.0870 & 0.1059 & & ne ii & 3367.22 & v20 & 3p 2d * & 3d 2f & 6 & 8 & 11.49 + & & & * & ne ii & 3366.98 & v12 & 3p 4d * & 3d 4d & 6 & 8 & + & & & * & n iii & 3367.36 & v5 & 3s 4p * & 3p 4p & 6 & 6 & + 3370.30 & 0.0490 & 0.0596 & & co ii & 3370.92 & v2 & a5p & z5d * & 3 & 1 & 20.41 + 3373.30 & 0.0120 & 0.0146 & & ne ii & 3374.06 & v12 & 3p 4d * & 3d 4d & 4 & 4 & 83.33 + 3373.67 & 0.0170 & 0.0207 & & ? & & & & & & & 58.82 + 3377.33 & 0.0240 & 0.0292 & & ? & & & & & & & 41.67 + & & & * & ne ii & 3379.32 & v12 & 3p 4d * & 3d 4d & 2 & 2 & + 3386.35 & 0.0450 & 0.0547 & & [ fe ii ] & 3387.09 & & a4f & b4d & 10 & 6 & 22.22 + & & & * & ne ii & 3386.20 & v12 & 3p 4d * & 3d 4d & 4 & 6 & + 3389.85 & 0.0140 & 0.0170 & & ne ii & 3390.55 & v12 & 3p 4d * & 3d 4d & 2 & 4 & 71.43 + 3392.12 & 0.0090 & 0.0109 & & ne ii & 3393.18 & v21 & 3p 2d * & 3d 2d & 6 & 4 & : + 3402.91 & 0.0490 & 0.0594 & & o iv & 3403.52 & v2 & 3p 2p * & 3d 2d & 2 & 4 & 20.41 + 3405.11 & 0.1630 & 0.1975 & & o iii & 3405.71 & v15 & 3p 3p & 3d 3p * & 1 & 3 & 6.13 + 3407.12 & 0.0860 & 0.1042 & & o iii & 3408.12 & v15 & 3p 3p & 3d 3p * & 3 & 1 & 11.63 + 3411.08 & 0.0400 & 0.0484 & & o iv & 3411.69 & v2 & 3p 2p * & 3d 2d & 4 & 6 & 25.00 + 3413.03 & 0.0450 & 0.0545 & & o iv & 3413.64 & v2 & 3p 2p * & 3d 2d & 4 & 4 & 22.22 + 3414.61 & 0.3430 & 0.4152 & & o iii & 3415.26 & v15 & 3p 3p & 3d 3p * & 3 & 3 & 2.92 + 3416.27 & 0.0640 & 0.0775 & & ne ii & 3416.91 & v21 & 3p 2d * & 3d 2d & 6 & 6 & 15.62 + 3417.05 & 0.0300 & 0.0363 & & ne ii & 3417.69 & v19 & 3p 2d * & 3d 4f & 6 & 8 & 33.33 + 3422.86 & 0.0770 & 0.0931 & & ne i & 3423.91 & v2.4 & 3s 2@xmath402@xmath41 * & 4p 1@xmath402@xmath41 & 3 & 3 & 12.99 + 3427.92 & 1.1660 & 1.4091 & & o iii & 3428.62 & v15 & 3p 3p & 3d 3p * & 3 & 5 & 0.86 + & & & * & o ii & 3428.51 & & 3d 4p & 5p 4s * & 6 & 4 & + 3429.87 & 0.2640 & 0.3190 & & o iii & 3430.57 & v15 & 3p 3p & 3d 3p * & 5 & 3 & 3.79 + & & & * & he i & 3430.56 & & 2p 3p * & 35s 3s & 1 & 3 & + 3433.94 & 0.0170 & 0.0205 & & he i & 3434.68 & & 2p 3p * & 31s 3s & & & 58.82 + 3443.33 & 9.5950 & 11.5747 & & o iii & 3444.06 & v15 & 3p 3p & 3d 3p * & 5 & 5 & 2.57 + 3446.76 & 0.1680 & 0.2026 & & he i & 3447.59 & v7 & 2s 1s & 6p 1p * & 1 & 3 & 2.86 + 3449.81 & 0.0180 & 0.0217 & & he i & 3450.28 & & 2p 3p * & 21d 3d & 9 & 15 & 55.56 + & & & * & o iii & 3450.92 & v25 & 3p 5d * & 3d 5f & 7 & 9 & + 3452.69 & 0.0160 & 0.0193 & & he i & 3453.32 & & 2p 3p * & 20d 3d & 9 & 15 & 62.50 + & & & * & ne ii & 3453.07 & v21 & 3p 2d * & 3d 2d & 4 & 4 & + 3456.09 & 0.0295 & 0.0355 & & he i & 3456.86 & & 2p 3p * & 19d 3d & 9 & 15 & 33.33 + & & & * & ne ii & 3456.61 & v28 & 3p 2s * & 3d 2p & 2 & 4 & + 3459.92 & 0.0320 & 0.0385 & & he i & 3461.01 & & 2p 3p * & 18d 3d & 9 & 15 & 31.25 + 3463.15 & 0.0140 & 0.0168 & & n iii & 3463.74 & & 2d & 2f * & 6 & 6 & 71.43 + 3465.25 & 0.0360 & 0.0433 & & he i & 3465.94 & v45 & 2p 3p * & 17d 2d & 9 & 15 & 27.78 + & & & * & o iii & 3466.11 & v25 & 3p 5d * & 3d 5f & 9 & 9 & + 3470.97 & 0.0510 & 0.0613 & & he i & 3471.81 & v44 & 2p 3p * & 16d 3d & 9 & 15 & 19.61 + 3476.93 & 0.0100 & 0.0120 & & ne ii & 3477.65 & v21 & 3p 2d * & 3d 2d & 4 & 6 & 30.00 + & & & * & n iv & 3478.71 & v1 & 3s 3s & 3p 3p * & 3 & 5 & + 3478.27 & 0.0730 & 0.0877 & & he i & 3478.97 & v43 & 2p 3p * & 15d 3d & 9 & 15 & 13.70 + 3479.78 & 0.0450 & 0.0541 & & o ii & 3480.25 & & 3d 4p & 5p 4p * & 4 & 4 & 22.22 + & & & * & [ fe iii ] & 3480.57 & & 3f4 & 3f2 & 9 & 5 & + 3481.21 & 0.0362 & 0.0434 & & ne ii & 3481.93 & v6 & 3s 2p & 3p 2s * & 4 & 2 & 33.33 + & & & * & [ fe iii ] & 3481.82 & & 3f4 & 3p2 & 7 & 5 & + & & & * & n iv & 3482.99 & v1 & 3s 3s & 3p 3p * & 3 & 3 & + 3486.99 & 0.0850 & 0.1020 & & he i & 3487.72 & v42 & 2p 3p * & 14d 3d & 9 & 15 & 9.41 + 3495.16 & 0.0180 & 0.0216 & & [ fe iii ] & 3496.08 & & 3f4 & 3f2 & 7 & 7 & 27.78 + 3497.82 & 0.1030 & 0.1235 & & he i & 3498.64 & v40 & 2p 3p * & 13d 3d & 9 & 15 & 7.77 + 3502.42 & 0.0100 & 0.0120 & & ne ii & 3503.58 & v28 & 3p 2s * & 3d 2p & 2 & 2 & 50.00 + 3511.64 & 0.1340 & 0.1604 & & he i & 3512.51 & v38 & 2p 3p * & 12d 3d & 9 & 15 & 5.97 + 3526.70 & 0.0100 & 0.0120 & & ne ii & 3527.36 & & 2f * & 2f & 8 & 8 & 50.00 + 3529.70 & 0.1590 & 0.1899 & & he i & 3530.49 & v36 & 2p 3p * & 11d 3d & 9 & 15 & 3.14 + 3536.21 & 0.0190 & 0.0227 & & he i & 3536.94 & & 2p 3p * & 11s 3s & 9 & 3 & 26.32 + & & & * & ne ii & 3537.00 & & 3d 2d & 5p 2p * & 4 & 4 & + 3539.65 & 0.0110 & 0.0131 & & ni i & 3540.33 & v119 & z5f * & i3f & 3 & 5 & 45.45 + 3542.01 & 0.0260 & 0.0310 & & ne ii & 3542.85 & v34 & 3p 4s * & 3d 4p & 4 & 6 & 19.23 + 3544.56 & 0.0160 & 0.0191 & & ? & & & & & & & 31.25 + 3549.78 & 0.0160 & 0.0191 & & si iii & 3550.38 & v8.01 & 3d 3f * & 5d 3d & 5 & 3 & 31.25 + 3553.60 & 0.2220 & 0.2645 & & he i & 3554.41 & v34 & 2p 3p * & 10d 3d & 9 & 15 & 2.25 + 3557.00 & 0.0120 & 0.0143 & & ne ii & 3557.81 & v6 & 3s 2p & 3p 2s * & 2 & 2 & 41.67 + 3562.01 & 0.0240 & 0.0286 & & he i & 3562.95 & v33 & 2p 3p * & 10s 3s & 9 & 3 & 22.50 + 3564.83 & 0.0210 & 0.0250 & & ne ii & 3565.82 & v34 & 3p 4s * & 3d 4p & 4 & 4 & 26.67 + 3567.62 & 0.1410 & 0.1678 & & ne ii & 3568.50 & v9 & 3s 2d & 3p 2f * & 6 & 8 & 9.93 + 3573.68 & 0.0450 & 0.0535 & & ne ii & 3574.18 & v9 & 3s 2d & 3p 2f * & 6 & 6 & 4.67 + & & & * & ne ii & 3574.61 & v9 & 3s 2d & 3p 2f * & 4 & 6 & + 3577.71 & 0.0050 & 0.0059 & & fe i & 3578.55 & & a3 g & n7p * & 7 & 9 & : + 3580.06 & 0.0050 & 0.0059 & & c iii & 3580.68 & & 3 g & 3 g * & 7 & 11 & : + 3586.43 & 0.2920 & 0.3467 & & he i & 3587.27 & v31 & 2p 3p * & 9d 3d & 9 & 15 & 1.71 + 3589.55 & 0.0320 & 0.0380 & & ne ii & 3590.45 & v31 & 3p 4s * & 3d 4f & 4 & 6 & 15.62 + & & & * & c i & 3590.21 & v7.08 & 3s 3p * & 7p 3p & 3 & 5 & + 3593.58 & 0.0100 & 0.0119 & & ne ii & 3594.16 & v34 & 3p 4s * & 3d 4p & 4 & 2 & 50.00 + 3598.29 & 0.0200 & 0.0237 & & he i & 3599.32 & v30 & 2p 3p * & 9s 3s & 9 & 3 & 14.50 + 3600.16 & 0.0220 & 0.0261 & & ? & & & & & & & 13.64 + 3602.58 & 0.0120 & 0.0142 & & ne ii & 3603.24 & & 3d 4d & 5p 4p * & 4 & 4 & 18.33 + 3612.77 & 0.2440 & 0.2890 & & he
i & 3613.64 & v6 & 2s 1s & 5p 1p * & 1 & 3 & 4.30 + 3633.36 & 0.4530 & 0.5354 & & he i & 3634.23 & v28 & 2p 3p * & 8d 3d & 9 & 15 & 3.18 + 3643.10 & 0.0160 & 0.0189 & & ne ii & 3643.93 & v5 & 3s 2p & 3p 2d * & 4 & 4 & 12.50 + 3647.12 & 0.0180 & 0.0212 & & [ fe ii ] & 3647.78 & & a4d & b2d & 2 & 6 & 11.11 + 3650.87 & 0.0140 & 0.0165 & & h i & 3651.82 & h50 & 2p+ 2p * & 50d+ 2d & 8 & & 35.71 + & & & * & he i & 3652.00 & v27 & 2p 3p * & 8s 3s & 9 & 3 & + 3663.65 & 0.0760 & 0.0893 &
& h i & 3664.68 & h28 & 2p+ 2p * & 28d+ 2d & 8 & & 6.58 + & & & * & ne ii & 3664.07 & v1 & 3s 4p & 3p 4p * & 6 & 4 & + 3665.24 & 0.1360 & 0.1598 &
& h i & 3666.09 & h27 & 2p+ 2p * & 27d+ 2d & 8 & & 7.35 + 3666.89 & 0.1930 & 0.2267 &
& h i & 3667.68 & h26 & 2p+ 2p * & 26d+ 2d & 8 & & 5.18 + 3668.63 & 0.2210 & 0.2596 & & h i & 3669.46 & h25 & 2p+ 2p * & 25d+ 2d & 8 & & 4.52 + 3670.59 & 0.3530 & 0.4144 & & h i & 3671.48 & h24 & 2p+ 2p * & 24d+ 2d & 8 & & 40.23 + 3672.95 & 0.4180 & 0.4905 & & h i & 3673.76 & h23 & 2p+ 2p * & 23d+ 2d & 8 & & 34.86 + 3675.50 & 0.5050 & 0.5924 & & h i & 3676.36 & h22 & 2p+ 2p * & 22d+ 2d & 8 & & 16.22 + 3678.51 & 0.5570 & 0.6530 & & h i & 3679.35 & h21 & 2p+ 2p * & 21d+ 2d & 8 & & 11.92 + 3681.97 & 0.6220 & 0.7287 & & h i & 3682.81 & h20 & 2p+ 2p * & 20d+ 2d & 8 & & 5.98 + 3685.99 & 0.7120 & 0.8336 & & h i & 3686.83 & h19 & 2p+ 2p * & 19d+ 2d & 8 & & 1.10 + 3690.71 & 0.7780 & 0.9100 & & h i & 3691.55 & h18 & 2p+ 2p * & 18d+ 2d & 8 & & 5.27 + 3693.42 & 0.2170 & 0.2537 & & ne ii & 3694.21 & v1 & 3s 4p & 3p 4p * & 6 & 6 & 10.14 + 3696.29 & 0.9310 & 1.0879 & & h i & 3697.15 & h17 & 2p+ 2p * & 17d+ 2d & 8 & & 4.68 + 3701.08 & 0.2550 & 0.2977 & & ne ii & 3701.78 & v40 & 3p 2p * & 3d 4p & 4 & 6 & 9.96 + & & & * & o iii & 3702.75 & v14 & 3p 3p & 3d 3d * & 1 & 3 & + 3702.90 & 1.3020 & 1.5196 &
& h i & 3703.85 & h16 & 2p+ 2p * & 16d+ 2d & 8 & & 2.91 + 3704.20 & 0.7640 & 0.8915 & & he i & 3705.00 & v25 & 2p 3p * & 7d 3d & 9 & 15 & 6.06 + 3706.49 & 0.2860 & 0.3336 & & o iii & 3707.25 & v14 & 3p 3p & 3d 3d * & 3 & 5 & 7.45 + & & & * & he i & 3707.34 & & 2p 1p * & 23s 1s & 3 & 1 & + 3708.82 & 0.0940 & 0.1096 & & ne ii & 3709.62 & v1 & 3s 4p & 3p 4p * & 4 & 2 & 23.09 + & & & * & he i & 3709.65 & & 2p 1p * & 22d 1d & 3 & 5 & + 3711.17 & 1.3790 & 1.6071 &
& h i & 3711.97 & h15 & 2p+ 2p * & 15d+ 2d & 8 & & 2.89 + & & & * & he i & 3712.70 & & 2p 1p * & 21d 1d & 3 & 5 & + & & & * & o ii & 3712.74 & v3 & 3s 4p & 3p 4s * & 4 & 4 & + 3712.28 & 0.2630 & 0.3064 & & ne ii & 3713.08 & v5 & 3s 2p & 3p 2d * & 4 & 6 & 9.51 + & & & * & o iii & 3714.03 & v14 & 3p 3p & 3d 3d * & 3 & 3 & + 3714.32 & 0.2350 & 0.2737 & & o iii & 3715.08 & v14 & 3p 3p & 3d 3d * & 5 & 7 & 13.74 + 3714.40 & 0.0210 & 0.0245 & & he ii & 3715.16 & 4.29 & 4f+ 2f * & 29g+ 2 g & 32 & & 47.62 + & & & * & he i & 3716.22 & & 2p 1p * & 20d 1d & 3 & 5 & + & & & * & cr ii & 3715.19 & v20 & b4d & z4f * & 8 & 10 & + 3719.70 & 0.0150 & 0.0175 & & he ii & 3720.41 & 4.28 & 4f+ 2f * & 28g+ 2 g & 32 & & 33.33 + 3720.98 & 0.9280 & 1.0796 & & [ s iii ] & 3721.69 & f2 & 3p2 3p & 3p2 1s & 3 & 1 & 2.16 + 3721.23 & 2.5410 & 2.9560 & & h i & 3721.94 & h14 & 2p+ 2p * & 14d+ 2d & 8 & & 2.22 + & & & * & o iii & 3725.31 & v14 & 3p 3p & 3d 3d * & 5 & 5 & + 3725.17 & 11.8090 & 13.7280 & & [ o ii ] & 3726.03 & f1 & 2p3 4s * & 2p3 2d * & 4 & 4 & 0.17 + 3725.56 & 0.0170 & 0.0198 & & he ii & 3726.27 & 4.27 & 4f+ 2f * &
27g+ 2 g & 32 & & 29.41 + & & & * & ne ii & 3727.11 & v5 & 3s 2p & 3p 2d * & 2 & 4 & + 3727.92 & 5.9400 & 6.9019 & & [ o ii ] & 3728.81 & f1 & 2p3 4s * & 2p3 2d * & 4 & 6 & 0.34 + & & & * & o iii & 3732.12 & v14 & 3p 3p & 3d 3d * & 5 & 3 & + 3732.12 & 0.0180 & 0.0209 & & he ii & 3732.83 & 4.26 & 4f+ 2f * & 26g+ 2 g & 32 & & 27.78 + 3732.17 & 0.0290 & 0.0337 & & he i & 3732.88 & v24 & 2p 3p * & 7s 3s & 9 & 3 & 17.24 + 3733.55 & 1.9220 & 2.2310 & & h i & 3734.37 & h13 & 2p+ 2p * & 13d+ 2d & 8 & & 1.75 + & & & * & ne ii & 3734.94 & v1 & 3s 4p & 3p 4p * & 4 & 4 & + 3739.32 & 0.0450 & 0.0522 & & o ii & 3739.76 & v31 & 3p 4s * & 4s 4p & 4 & 6 & 11.11 + 3739.52 & 0.0210 & 0.0244 & & he ii & 3740.23 & 4.25 & 4f+ 2f * & 25g+ 2 g & 32 & & 47.62 + 3747.78 & 0.0210 & 0.0243 & & he ii & 3748.61 & 4.24 & 4f+ 2f * & 24g+ 2 g & 32 & & 23.81 + 3749.32 & 2.4810 & 2.8719 & & h i & 3750.15 & h12 & 2p+ 2p * & 12d+ 2d & 8 & & 2.55 + & & & * & ne ii & 3751.25 & v1 & 3s 4p & 3p 4p * & 2 & 2 & + 3753.94 & 0.5688 & 0.6579 & & o iii & 3754.70 & v2 & 3s 3p * & 3p 3d & 3 & 5 & 5.46 + & & & * & he i & 3756.10 & v66 & 2p 1p * & 14d 1d & 3 & 5 & + 3756.43 & 0.3091 & 0.3574 & & o iii & 3757.24 & v2 & 3s 3p * & 3p 3d & 1 & 3 & 20.13 + & & & * & he ii & 3758.15 & 4.23 & 4f+ 2f * & 23g+ 2 g & 32 & & 41.67 + 3759.09 & 1.5782 & 1.8238 & & o iii & 3759.87 & v2 & 3s 3p * & 3p 3d & 5 & 7 & 2.97 + 3761.36 & 0.0810 & 0.0936 & & mg i & 3762.10 & & 1d & 1p * & 5 & 3 & 12.35 + & & & * & ne ii & 3766.26 & v1 & 3s 4p & 3p 4p * & 4 & 6 & + 3768.02 & 0.0280 &
0.0323 & & he i & 3768.81 & v65 & 2p 1p * & 13d 1d & 3 & 5 & 2.14 + 3768.29 & 0.0260 & 0.0300 & & he ii & 3769.08 & 4.22 & 4f+ 2f * & 22g+ 2 g & 32 & & 2.31 + 3769.84 & 3.2770 & 3.7797 & & h i & 3770.63 & h11 & 2p+ 2p * & 11d+ 2d & 8 & & 2.17 + 3773.12 & 0.1670 & 0.1925 & &
o iii & 3774.02 & v2 & 3s 3p * & 3p 3d & 3 & 3 & 32.16 + 3776.37 & 0.0430 & 0.0495 & & ne ii & 3777.14 & v1 & 3s 4p & 3p 4p * & 2 & 4 & 23.26 + 3780.89 & 0.0380 & 0.0437 & & he ii & 3781.68 & 4.21 & 4f+ 2f * & 21g+ 2 g & 32 & & 26.32 + 3783.92 & 0.0320 & 0.0368 & & he i & 3784.86 & v64 & 2p 1p * & 12d 1d & 3 & 5 & 31.25 + 3790.52 & 0.1554 & 0.1786 & & o iii & 3791.28 & v2 & 3s 3p * & 3p 3d & 5 & 5 & 20.62 + 3795.55 & 0.4280 & 0.4915 & & he ii & 3796.33 & 4.20 & 4f+ 2f * & 20g+ 2 g & 32 & & 11.12 + 3797.12 & 4.3400 & 4.9824 & & h i & 3797.90 & h10 & 2p+ 2p * & 10d+ 2d & 8 & & 1.63 + 3801.49 & 0.0050 & 0.0057 & & ne ii & 3802.29 & & 4d 4d & 4f2@xmath401@xmath41 * & 6 & 4 & : + 3805.00 & 0.0430 & 0.0493 & & he i & 3805.74 & v63 & 2p 1p * & 11d 1d & 3 & 5 & 23.26 + 3805.82 & 0.0200 & 0.0229 & & si iii & 3806.54 & v5 & 4p 3p * & 4d 3d & 10 & 12 & 50.00 + & & & * &
si iii & 3806.78 & v5 & 4p 3p * & 4d 3d & 5 & 3 & + 3810.24 & 0.0100 & 0.0115 & & o iii & 3810.99 & v2 & 3s 3p * & 3p 3d & 5 & 3 & : + 3812.74 & 0.0570 & 0.0653 & & he ii & 3813.49 & 4.19 & 4f+ 2f * & 19g+ 2 g & 32 & & 35.09 + 3818.86 & 1.1470 & 1.3119 & & he i & 3819.60 & v22 & 2p 3p * & 6d 3d & 9 & 15 & 4.84 + 3832.80 & 0.0640 & 0.0730 & & he i & 3833.57 & v62 & 2p 1p * & 10d 1d & 3 & 5 & 0.94 + 3833.03 & 0.0460 & 0.0525 & & he ii & 3833.80 & 4.18 & 4f+ 2f * & 18g+ 2 g & 32 & & 0.87 + 3834.61 & 6.6300 & 7.5634 & & h i & 3835.38 & h9 & 2p+ 2p * & 9d+ 2d & 8 & & 0.93 + 3837.60 & 0.0020 & 0.0023 & & n ii & 3838.37 & v30 & 3p 3p & 4s 3p * & 5 & 5 & : + 3838.63 & 0.0850 & 0.0969 & & [ fe v ] & 3839.27 & & 3d4 5d & 3d4 3f2 & 7 & 7 & 0.94 + 3841.41 & 0.0004 & 0.0005 & & n ii & 3842.18 & v30 & 3p 3p & 4s 3p * & 1 & 3 & : + 3842.08 & 0.0140 & 0.0160 & & o ii & 3842.82 & v12 & 3p 4d * & 3d 4d & 2 & 4 & 5.71 + 3847.16 & 0.0120 & 0.0137 & & o ii & 3847.89 & v12 & 3p 4d * & 3d 4d & 2 & 2 & 5.00 + 3850.06 & 0.0030 & 0.0034 & & o ii & 3850.80 & v12 & 3p 4d * & 3d 4d & 4 & 6 & 6.67 + 3850.29 & 0.0150 & 0.0171 & & o ii & 3851.03 & v12 & 3p 4d * & 3d 4d & 4 & 4 & 6.00 + 3850.61 & 0.0070 & 0.0080 & & [ fe v ] & 3851.35 & & 3d4 5d & 3d4 3f2 & 7 & 5 & 65.71 + 3852.93 & 0.0070 & 0.0080 & & si ii & 3853.66 & v1 & 3p2 2d & 4p 2p * & 4 & 4 & 37.14 + 3855.28 & 0.0461 & 0.0524 & & si ii & 3856.02 & v1 & 3p2 2d & 4p 2p * & 6 & 4 & 7.45 + 3855.39 & 0.0119 & 0.0135 & & o ii & 3856.13 & v12 & 3p 4d * & 3d 4d & 4 & 2 & 6.36 + 3857.40 & 0.0740 & 0.0841 & & he ii & 3858.07 & 4.17 & 4f+ 2f * & 17g+ 2 g & 32 & & 5.41 + 3862.18 & 0.0610 & 0.0693 & & si ii & 3862.60 & v1 & 3p2 2d & 4p 2p * & 4 & 2 & 8.20 + & & & * & ne
ii & 3862.59 & & 5 g 2@xmath404@xmath41 & 5f2@xmath404@xmath41 * & 8 & 10 & + & & & * & o ii & 3863.50 & v12 & 3p 4d * & 3d 4d & 6 & 8 & + & & & * & o ii & 3864.43 & v12 & 3p 4d * & 3d 4d & 6 & 6 & + & & & * & o ii & 3864.67 & v12 & 3p 4d * & 3d 4d & 6 & 4 & + 3867.96 & 104.4480 & 118.4975 & & [ ne iii ] & 3868.76 & f1 & 2p4 3p & 2p4 1d & 5 & 5 & 0.01 + 3877.30 & 0.0300 & 0.0340 & & he i & 3878.18 & v59 & 2p 1p * & 9s 1s & 3 & 1 & 6.67 + 3881.45 & 0.0470 & 0.0532 & & o ii & 3882.19 & v12 & 3p 4d * & 3d 4d & 8 & 8 & 1.28 + 3881.51 & 0.0370 & 0.0419 & & ? & & & & & & & 1.08 + 3881.71 & 0.0050 & 0.0057 & & o ii & 3882.45 & v11 & 3p 4d * & 3d 4p & 4 & 4 & 2.00 + 3882.39 & 0.0030 & 0.0034 & & o ii & 3883.13 & v12 & 3p 4d * & 3d 4d & 8 & 6 & 3.33 + 3883.08 & 0.0290 & 0.0328 & & c iii & 3883.82 & v15 & 4d 3d & 5f 3f * & 3 & 5 & 1.03 + & & & * & c ii & 3883.82 & v33 & 3d 4f * & 4f 4 g & 10 & 8 & + 3886.70 & 0.0640 & 0.0724 & & he ii & 3887.44 & 4.16 & 4f+ 2f * & 16g+ 2 g & 32 & & 1.25 + 3887.91 & 9.9670 & 11.2710 & & he i & 3888.64 & v2 & 2s 3s & 3p 3p * & 3 & 9 & 1.19 + 3888.31 & 8.5920 & 9.7155 & & h i & 3889.05 & h8 & 2p+ 2p * & 8d+ 2d & 8 & & 1.19 + 3890.58 & 0.1620 & 0.1831 & & [ fe v ] & 3891.28 & & 3d4 5d & 3d4 3f2 & 9 & 9 & 6.17 + 3894.52 & 0.0490 & 0.0554 & & [ fe v ] & 3895.22 & & 3d4 5d & 3d4 3p2 & 7 & 5 & 20.41 + 3895.26 & 0.0390 & 0.0441 & & fe ii ] & 3896.10 & v23 & a2d & z4p * & 4 & 6 & 25.64 + 3901.99 & 0.0100 & 0.0113 & & p ii & 3902.87 & & 3d 3d * & 5p 3d & 7 & 5 & 50.00 + & & & * & fe iii & 3902.81 & & 4 g 4d5f & 4d 5p5f * & 11 & 11 & + 3906.61 & 0.0090 & 0.0101 & & o ii & 3907.45 & v11 & 3p 4d * & 3d 4p & 6 & 6 & 55.56 + 3906.90 & 0.0080 & 0.0090 & & [ cr iii ] & 3907.74 & & b1 g & 1h & 9 & 11 & 62.50 + 3911.06 & 0.0200 & 0.0225 & & o ii & 3911.96 & v17 & 3s 2d & 3p 2p * & 6 & 4 & 25.00 + 3914.16 & 0.0030 & 0.0034 & & c iii & 3914.90 & & 3p * & 3d & 5 & 3 & : + 3918.19 & 0.0140 & 0.0158 & & c ii & 3918.98 & v4 & 3p 2p * & 4s 2s & 2 & 2 & 19.29 + 3918.50 & 0.0100 & 0.0113 & & o ii & 3919.29 & v17 & 3s 2d & 3p 2p * & 4 & 2 & 28.00 + & & & * & n ii & 3919.00 & v17 & 3p 1p & 3d 1p * & 3 & 3 & + 3919.90 & 0.0300 & 0.0338 & & c ii & 3920.69 & v4 & 3p 2p * & 4s 2s & 4 & 2 & 3.33 + 3922.74 & 0.1060 & 0.1192 & & he ii & 3923.48 & 4.15 & 4f+ 2f * & 15g+ 2 g & 32 & & 2.55 + 3923.64 & 0.0120 & 0.0135 & & si iii & 3924.47 & v8.14 & 4f 1f * & 5
g 1 g & 7 & 9 & 15.83 + 3925.79 & 0.1170 & 0.1315 & & he i & 3926.53 & v58 & 2p 1p * & 8d 1d & 3 & 5 & 0.94 + 3925.81 & 0.0020 & 0.0022 & & o ii & 3926.58 & v11 & 3p 4d * & 3d 4p & 8 & 6 & 55.00 + 3928.01 & 0.0130 & 0.0146 & & c ii & 3928.80 & & 4p * & 4d & 2 & 2 & 6.92 + 3930.87 & 0.0050 & 0.0056 & & o ii & 3931.55 & & g@xmath405@xmath41 * & 2@xmath406@xmath41 & 10 & 12 & 28.00 + 3935.11 & 0.0100 & 0.0112 & & he i & 3935.91 & v57 & 2p 1p * & 8s 1s & 3 & 1 & 20.00 + 3941.45 & 0.0060 & 0.0067 & & ar ii & 3942.28 & & 3d `` 2d & 5f 2@xmath404@xmath41 * & 6 & 8 & 33.33 + 3944.18 & 0.0230 & 0.0258 & & o ii & 3945.04 & v6 & 3s 2p & 3p 2p * & 2 & 4 & 7.20 + & & & * & c ii & 3945.00 & v32 & 3d 4f * & 4f 4f & 4 & 4 & + & & & * & c ii & 3945.20 & v32 & 3d 4f * & 4f 4f & 6 & 8 & + 3947.28 & 0.0270 & 0.0302 & & o iii & 3948.01 & v26.04 & 4p 3d & 5s 3p * & 3 & 5 & 7.41 + & & & * & c ii & 3947.72 & v32 & 3d 4f * & 4f 4f & 8 & 10 & + & & & * & c ii & 3948.33 & v32 & 3d 4f * & 4f 4f & 8 & 8 & + 3953.72 & 0.0270 & 0.0302 & & o ii & 3954.36 & v6 & 3s 2p & 3p 2p * & 2 & 2 & 7.41 + 3960.77 & 0.0280 & 0.0313 & & o iii & 3961.57 & v17 & 3p 1d & 3d 1f * & 5 & 7 & 17.86 + 3963.93 & 1.0640 & 1.1888 & & he i & 3964.73 & v5 & 2s 1s & 4p 1p * & 1 & 3 & 8.18 + 3966.67 & 30.4230 & 33.9772 & & [ ne iii ] & 3967.47 & f1 & 2p4 3p & 2p4 1d & 3 & 5 & 0.46 + 3967.63 & 0.0970 & 0.1083 & & he ii & 3968.43 & 4.14 & 4f+ 2f * & 14g+ 2 g & 32 & & 5.15 + 3969.27 & 14.6370 & 16.3404 & & h i & 3970.07 & h7 & 2p+ 2p * & 7d+ 2d & 8 & 98 & 0.32 + & & & * & o iii & 3971.33 & v43 & 4d 3p * & 3d 3d & 5 & 7 & + 3972.49 & 0.0580 & 0.0647 & & o ii & 3973.26 & v6 & 3s 2p & 3p 2p * & 4 & 4 & 8.62 + & & & * & o iii & 3979.12 & v43 & 4d 3p * & 3d 3d & 5 & 5 & + & & & * & o iii & 3980.91 & v43 & 4d 3p * & 3d 3d & 3 & 5 & + 3981.93 & 0.0090 & 0.0100 & & o ii & 3982.71 & v6 & 3s 2p & 3p 2p * & 4 & 2 & 55.56 + 3982.90 & 0.0110 & 0.0123 & & o iii & 3983.73 & v43 & 4d 3p * & 3d 3d & 5 & 3 & 45.45 + 3985.05 & 0.0150 & 0.0167 & & n ii & 3985.85 & & ? & ? & & & 33.33 + & & & * & o iii & 3985.55 & v43 & 4d 3p * & 3d 3d & 3 & 3 & + 3994.26 & 0.0300 & 0.0334 & & n ii & 3994.99 & v12 & 3s 1p * & 3p 1d & 3 & 5 & 6.00 + & & & * & fe ii & 3995.13 & & y4d & 5d 4d4p & 4 & 6 & + 3997.95 & 0.0226 & 0.0252 & & n iii & 3998.63 & v17 & 4d 2d & 5f 2f * & 4 & 6 & 9.29 + 4002.76 & 0.0349 & 0.0389 & & n iii & 4003.58 & v17 & 4d 2d & 5f 2f * & 6 & 8 & 5.73 + & & & * & n iii & 4003.72 & v17 & 4d 2d & 5f 2f * & 6 & 6 & + 4007.63 & 0.0160 & 0.0178 & & [ fe iii ] & 4008.35 & & 3d6 5d & 3d6 3 g & 9 & 9 & 11.25 + 4008.43 & 0.1930 & 0.2142 & & he i & 4009.26 & v55 & 2p 1p * & 7d 1d & 3 & 5 & 1.66 + 4011.23 & 0.0130 & 0.0144 & & [ ne iii ] & 4012.01 & f1 & 2p4 3p & 2p4 1d & 1 & 5 & 13.85 + 4015.20 & 0.0060 & 0.0067 & & fe iii & 4015.99 & & 4p 4d5f & 6s 5f5f * & 11 & 11 & 30.00 + 4023.14 & 0.0230 & 0.0255 & & he
i & 4023.99 & v54 & 2p 1p * & 7s 1s & 3 & 1 & : + & & & * & o ii & 4023.86 & v99 & 3d 2f & 4f d2 * & 8 & 6 & + 4024.75 & 0.1250 & 0.1384 & & he ii & 4025.60 & 4.13 & 4f+ 2f * & 13g+ 2 g & 32 & & : + 4025.23 & 0.0160 & 0.0177 & & n ii & 4026.08 & v39b & 3d 3f * & 4f 2@xmath405@xmath41 & 7 & 9 & : + 4025.35 & 2.1750 & 2.4075 & & he i & 4026.20 & v18 & 2p 3p * & 5d 3d & 9 & 15 & 1.26 + 4025.57 & 0.0020 & 0.0022 & & o ii & 4026.31 & v50b & 3d 4f & 4f f3 * & 4 & 6 & : + 4031.36 & 0.0050 & 0.0055 & & o ii & 4032.25 & v50a & 3d 4f & 4f f4 * & 6 & 8 & 2.00 + 4031.59 & 0.0006 & 0.0007 & & o ii & 4032.48 & v50c & 3d 4f & 4f f2 * & 4 & 6 & : + 4032.27 & 0.0110 & 0.0122 & & o ii & 4033.16 & v50c & 3d 4f & 4f f2 * & 4 & 4 & 1.82 + & & & * & [ ni ii ] & 4033.04 & & 3d9 2d & ( 3p)4s 4p & 6 & 4 & + & & & * & o ii & 4035.07 & v50b & 3d 4f & 4f f3 * & 6 & 6 & + 4034.19 & 0.0336 & 0.0371 & & n ii & 4035.08 & v39a & 3d 3f * & 4f 2@xmath404@xmath41 & 5 & 7 & 19.13 + & & & * & o ii & 4035.46 & v50b & 3d 4f & 4f f3 * & 6 & 8 & + 4038.65 & 0.0040 & 0.0044 & & n ii & 4039.35 & v39b & 3d 3f * & 4f 2@xmath405@xmath41 & 9 & 9 & 37.50 + 4040.58 & 0.0058 & 0.0065 & & o ii & 4041.28 & v50c & 3d 4f & 4f f2 * & 6 & 6 & 13.08 + 4040.60 & 0.0742 & 0.0819 & & n ii & 4041.31 & v39b & 3d 3f * & 4f 2@xmath405@xmath41 & 9 & 11 & 6.27 + & & & * & o ii & 4041.95 & v50c & 3d 4f & 4f f2 * & 6 & 4 & + 4042.83 & 0.0320 & 0.0354 & & n ii & 4043.53 & v39a & 3d 3f * & 4f 2@xmath404@xmath41 & 7 & 9 & 6.21 + 4044.07 & 0.0050 & 0.0055 & & n ii & 4044.78 & v39a & 3d 3f * & 4f 2@xmath404@xmath41 & 7 & 7 & 58.00 + 4044.23 & 0.0040 & 0.0044 & & o ii & 4044.95 & v50a & 3d 4f & 4f f4 * & 8 & 8 & 2.50 + 4045.40 & 0.0050 & 0.0055 & & o ii & 4046.12 & v50a & 3d 4f & 4f f4 * & 8 & 10 & 2.00 + 4047.51 & 0.0183 & 0.0202 & & o ii & 4048.22 & v50b & 3d 4f & 4f f3 * & 8 & 8 & 2.63 + & & & * & o ii & 4047.80 & v50b & 3d 4f & 4f f3 * & 8 & 6 & + 4051.65 & 0.0040 & 0.0044 & & [ mn v ] & 4052.28 & & 3d3 4f & 3d3 2d & 4 & 4 & 62.50 + 4053.58 & 0.0110 & 0.0121 & & o ii & 4054.52 & v98 & 3d 2f & 4f f3 * & 6 & 8 & 30.91 + & & & * & o ii & 4054.08 & v50c & 3d 4f & 4f f2 * & 8 & 6 & + & & & * & c iii & 4056.06 & v24 & 4d 1d & 5f 1f * & 5 & 7 & + 4056.40 & 0.0070 & 0.0077 & & n ii & 4056.90 & v39a & 3d 3f * & 4f 2@xmath404@xmath41 & 9 & 9 & 62.00 + 4057.66 & 0.0040 & 0.0044 & & n ii & 4058.16 & v39a & 3d 3f * & 4f 2@xmath404@xmath41 & 9 & 7 & 67.50 + 4059.19 & 0.0110 & 0.0121 & & [ f iv ] & 4059.90 & & 2p2 3p & 2p2 1d & 5 & 5 & 26.36 + 4059.89 & 0.0060 & 0.0066 & & o ii & 4060.60 & v97 & 3d 2f & 4f g4 * & 8 & 10 & 8.33 + & & & * & o ii & 4060.60 & v97 & 3d 2f & 4f g4 * & 8 & 8 & + 4060.31 & 0.0210 & 0.0231 & & o ii & 4061.03 & v97 & 3d 2f & 4f g4 * & 6 & 8 & 7.62 + 4060.49 & 0.0002 & 0.0002 & & o ii & 4061.21 & v46c & 3d 4f & 4f d1 * & 4 & 2 & : + & & & * & o ii & 4061.22 & v46c & 3d 4f & 4f d1 * & 4 & 4 & + 4061.03 & 0.0020 & 0.0022 & & o ii & 4061.76 & v50a & 3d 4f & 4f f4 * & 10 & 8 & : + 4062.22 & 0.0330 & 0.0363 & & o ii & 4062.94 & v50a & 3d 4f & 4f f4 * & 10 & 10 & 7.27 + 4062.25 & 0.0007 & 0.0008 & & ne ii & 4062.97 & v53 & 3d 4d & 4f 0@xmath403@xmath41 * & 6 & 8 & : + 4064.32 & 0.0030 & 0.0033 & & o ii & 4065.05 & v50b & 3d 4f & 4f f3 * & 10 & 8 & 6.67 + 4067.14 & 0.0650 & 0.0715 & & c iii & 4067.94 & v16 & 4f 3f * & 5
g 3 g & 5 & 7 & 39.69 + 4067.72 & 0.6610 & 0.7271 & & [ s ii ] & 4068.60 & f1 & 3p3 4s * & 3p3 2p * & 4 & 4 & 3.42 + 4068.12 & 0.0850 & 0.0935 & & c iii & 4068.92 & v16 & 4f 3f * & 5 g 3 g & 7 & 9 & 39.76 + & & & * & o ii & 4069.62 & v10 & 3p 4d * & 3d 4f & 2 & 4 & + 4069.09 & 0.5770 & 0.6346 & & o ii & 4069.89 & v10 & 3p 4d * & 3d 4f & 4 & 6 & : + 4069.50 & 0.1120 & 0.1232 & & c iii & 4070.31 & v16 & 4f 3f * & 5 g 3 g & 9 & 11 & 39.38 + & & & * & o ii & 4071.24 & v48a & 3d 4f & 4f g5 * & 8 & 10 & + & & & * & [ fe v ] & 4071.24 & & 3d4 5d & 3d4 3p2 & 5 & 3 & + 4071.37 & 0.4990 & 0.5486 & & o ii & 4072.16 & v10 & 3p 4d * & 3d 4f & 6 & 8 & 6.50 + & & & * & n ii & 4073.05 & v38b & 3d 3f * & 4f 1@xmath404@xmath41 & 5 & 7 & + 4073.47 & 0.0640 & 0.0703 & & o iii & 4074.02 & v23 & 3s 3p & 3p 3d * & 3 & 5 & 19.69 + & & & * & c ii & 4074.52 & v36 & 3d 4d * & 4f 4f & 2 & 4 & + & & & * & c ii & 4074.52 & v36 & 3d 4d * & 4f 4f & 4 & 6 & + & & & * & c ii & 4074.85 & v36 & 3d 4d * & 4f 4f & 6 & 8 & + 4075.11 & 0.6260 & 0.6878 & & o ii & 4075.86 & v10 & 3p 4d * & 3d 4f & 8 & 10 & 9.07 + & & & * & c ii & 4075.85 & v36 & 3d 4d * & 4f 4f & 8 & 10 & + 4075.47 & 0.2180 & 0.2395 & & [ s ii ] & 4076.35 & f1 & 3p3 4s * & 3p3 2p * & 4 & 2 & 3.44 + & & & * & c ii & 4076.53 & v36 & 3d 4d * & 4f 4f & 8 & 8 & + 4076.11 & 0.0100 & 0.0110 & & n ii & 4076.91 & v38a & 3d 3f * & 4f 1@xmath403@xmath41 & 5 & 5 & 2.00 + 4076.60 & 0.0001 & 0.0001 & & n ii & 4077.40 & v38a & 3d 3f * & 4f 1@xmath403@xmath41 & 7 & 7 & : + 4078.00 & 0.0810 & 0.0890 & & o ii & 4078.84 & v10 & 3p 4d * & 3d 4f & 4 & 4 & 10.49 + & & & * & [ fe iii ] & 4079.70 & & 5d 3d6 & 3 g 3d6 & 7 & 9 & + 4080.23 & 0.0370 & 0.0406 & & o iii & 4081.00 & v23 & 3s 3p & 3p 3d * & 5 & 7 & 19.46 + 4081.59 & 0.0260 & 0.0285 & & n ii & 4082.27 & v38b & 3d 3f * & 4f 1@xmath404@xmath41 & 7 & 9 & 2.31 + 4082.21 & 0.0020 & 0.0022 & & n ii & 4082.89 & v38b & 3d 3f * & 4f 1@xmath404@xmath41 & 7 & 7 & : + 4083.22 & 0.0787 & 0.0864 & & o ii & 4083.90 & v48b & 3d 4f & 4f g4 * & 6 & 8 & 5.12 + 4084.43 & 0.0975 & 0.1070 & & o ii & 4085.11 & v10 & 3p 4d * & 3d 4f & 6 & 6 & 6.79 + 4086.15 & 0.0010 & 0.0011 & & n ii & 4086.83 & v38a & 3d 3f * & 4f 1@xmath403@xmath41 & 7 & 5 & : + 4086.47 & 0.0840 & 0.0921 & & o ii & 4087.15 & v48c & 3d 4f & 4f g3 * & 4 & 6 & 7.14 + 4086.62 & 0.0060 & 0.0066 & & n ii & 4087.30 & v38a & 3d 3f * & 4f 1@xmath403@xmath41 & 7 & 7 & : + 4087.59 & 0.0050 & 0.0055 & & o ii & 4088.27 & v48a & 3d 4f & 4f g5 * & 10 & 10 & 2.00 + & & & * & cr ii & 4088.52 & & x6p * & 5d 6d6d & 4 & 6 & + 4088.17 & 0.0430 & 0.0472 & & si iv & 4088.86 & v1 & 4s 2s & 4p 2p * & 2 & 4 & 2.09 + 4088.53 & 0.2420 & 0.2654 & & o ii & 4089.29 & v48a & 3d 4f & 4f g5 * & 10 & 12 & 1.82 + 4092.19 & 0.0700 & 0.0767 & & o ii & 4092.93 & v10 & 3p 4d * & 3d 4f & 8 & 8 & 30.00 + 4093.40 & 0.0060 & 0.0066 & & o ii & 4094.14 & v10 & 3p 4d * & 3d 4f & 6 & 4 & 5.00 + 4094.90 & 0.0760 & 0.0833 & & o ii & 4095.64 & v48c & 3d 4f & 4f g3 * & 6 & 8 & 4.74 + 4095.16 & 0.0450 & 0.0493 & & n ii & 4095.90 & v38b & 3d 3f * & 4f 1@xmath404@xmath41 & 9 & 9 & 14.44 + & & & * & fe iii & 4095.84 & & 4 g 5p5 g * & 4 g 5d5 g & 11 & 11 & + 4095.45 & 0.0170 & 0.0186 & & o ii & 4096.19 & v48c & 3d 4f & 4f g3 * & 6 & 6 & 4.71 + 4095.84 & 0.0020 & 0.0022 & & n ii & 4096.58 & v38b & 3d 3f * & 4f 1@xmath404@xmath41 & 9 & 7 & : + & & & * & [ fe iii ] & 4096.61 & & 3d6 5d & 3d6 3 g & 5 & 7 & + 4096.19 & 0.0110 & 0.0120 & & o ii & 4096.93 & v48b & 3d 4f & 4f g4 * & 8 & 8 & 4.55 + 4096.51 & 0.0550 & 0.0602 & & o ii & 4097.25 & v20 & 3p 4p * & 3d 4d & 2 & 4 & 4.73 + 4096.52 & 0.2160 & 0.2366 & & o ii & 4097.26 & v48b & 3d 4f & 4f g4 * & 8 & 10 & 4.72 + 4096.55 & 1.6720 & 1.8315 & & n iii & 4097.33 & v1 & 3s 2s & 3p 2p * & 2 & 4 & 9.06 + 4096.64 & 0.0410 & 0.0449 & & n ii & 4097.38 & v44b & 3d 1d * & 4f 2@xmath402@xmath41 & 5 & 5 & 5.61 + & & & * & fe ii & 4097.50 & & y4d * & e4f & 8 & 10 & + 4097.27 & 0.0230 & 0.0252 & & n ii & 4098.01 & v44b & 3d 1d * & 4f 2@xmath402@xmath41 & 5 & 3 & 5.65 + 4100.20 & 0.0008 & 0.0009 & & n ii & 4100.97 & v38a & 3d 3f * & 4f 1@xmath403@xmath41 & 9 & 7 & 10.00 + 4100.94 & 25.7890 & 28.2308 & & h i & 4101.73 & h6 & 2p+ 2p * & 6d+ 2d & 8 & 72 & 2.09 + 4102.23 & 0.0490 & 0.0536 & & o ii & 4103.00 & v20 & 3p 4p * & 3d 4d & 2 & 2 & 1.84 + & & & * & n iii & 4103.39 & v1 & 3s 2s & 3p 2p * & 2 & 2 & + 4103.95 & 0.0170 & 0.0186 & & o ii & 4104.72 & v20 & 3p 4p * & 3d 4d & 4 & 6 & 1.76 + 4104.22 & 0.0950 & 0.1039 & & o ii & 4104.99 & v20 & 3p 4p * & 3d 4d & 4 & 4 & 1.79 + 4105.25 & 0.0040 & 0.0044 & & o ii & 4106.02 & v10 & 3p 4d * & 3d 4f & 8 & 6 & 2.50 + 4106.32 & 0.0530 & 0.0580 & & o ii & 4107.09 & v46a & 3d 4f & 4f d3 * & 6 & 8 & 2.20 + 4106.56 & 0.0030 & 0.0033 & & o ii & 4107.33 & v46a & 3d 4f & 4f d3 * & 6 & 6 & : + 4107.98 & 0.0100 & 0.0109 & & o ii & 4108.75 & v48c & 3d 4f & 4f g3 * & 8 & 8 & 2.00 + 4108.52 & 0.0002 & 0.0002 & & o ii & 4109.30 & v48c & 3d 4f & 4f g3 * & 8 & 6 & : + 4109.27 & 0.0310 & 0.0339 & & n ii & 4110.04 & v44a & 3d 1d * & 4f 2@xmath403@xmath41 & 5 & 5 & 5.81 + 4110.03 & 0.0230 & 0.0251 & & o ii & 4110.79 & v20 & 3p 4p * & 3d 4d & 4 & 2 & 1.74 + 4110.07 & 0.0002 & 0.0002 & & n ii & 4110.83 & v44a & 3d 1d * & 4f 2@xmath403@xmath41 & 5 & 7 & : + 4111.62 & 0.0110 & 0.0120 & & ne ii & 4112.40 & v54b & 3d 4d & 4f 1@xmath404@xmath41 * & 6 & 8 & : + & & & * & o ii & 4114.18 & v48b & 3d 4f & 4f g4 * & 10 & 8 & + & & & * & o ii & 4114.50 & v48b & 3d 4f & 4f g4 * & 10 & 10 & + 4115.36 & 0.0240 & 0.0262 & & si iv & 4116.10 & v1 & 4s 2s & 4p 2p * & 2 & 2 & 17.08 + 4118.48 & 0.1401 & 0.1530 & & o ii & 4119.22 & v20 & 3p 4p * & 3d 4d & 6 & 8 & 23.84 + 4119.54 & 0.0183 & 0.0200 & & o ii & 4120.28 & v20 & 3p 4p * & 3d 4d & 6 & 6 & : + 4119.80 & 0.0255 & 0.0279 & & o ii & 4120.54 & v20 & 3p 4p * & 3d 4d & 6 & 4 & : + 4120.05 & 0.2380 & 0.2598 & & he i & 4120.99 & v16 & 2p 3p * & 5s 3s & 9 & 3 & 4.37 + 4120.72 & 0.0150 & 0.0164 & & o ii & 4121.46 & v19 & 3p 4p * & 3d 4p & 2 & 2 & 42.00 + 4121.88 & 0.0120 & 0.0131 & & [ k v ] & 4122.62 & & 3p3 4s * & 3p3 2d * & 4 & 6 & 39.17 + 4127.41 & 0.0090 & 0.0098 & & si ii & 4128.07 & v3 & 3d 2d & 4f 2f * & 4 & 6 & 38.89 + & & & * & [ fe ii ] & 4128.75 & & ( 3p2)4s 4p & ( 5d)4p 4d * & 6 & 4 & + 4128.67 & 0.0220 & 0.0240 & & o ii & 4129.32 & v19 & 3p 4p * & 3d 4p & 4 & 2 & 13.18 + 4130.24 & 0.0020 & 0.0022 & & si ii & 4130.89 & v3 & 3d 2d & 4f 2f * & 6 & 8 & 5.00 + & & & * & si ii & 4130.87 & v3 & 3d 2d & 4f 2f * & 6 & 6 & + 4131.13 & 0.0033 & 0.0036 & & n ii & 4131.78 & v43.01 & 3d 1d * & 4f 2@xmath404@xmath41 & 5 & 7 & 23.18 + 4132.15 & 0.0720 & 0.0785 & & o ii & 4132.80 & v19 & 3p 4p * & 3d 4p & 2 & 4 & 8.37 + & & & * &
n ii & 4133.67 & v65 & 3s 5p & 3p 5s * & 5 & 5 & + 4135.48 & 0.0013 & 0.0014 & & n ii & 4136.39 & & ? & ? & & & : + 4136.97 & 0.0008 & 0.0009 & & o ii & 4137.72 & v46a & 3d 4f & 4f d3 * & 10 & 8 & : + 4138.99 & 0.0130 & 0.0142 & & o i & 4139.92 & & ? & ? & & & 40.77 + 4139.95 & 0.0030 & 0.0033 & & o ii & 4140.70 & v19 & 3p 4p * & 3d 4p & 4 & 4 & 43.33 + 4141.80 & 0.0560 & 0.0609 & & o ii & 4142.27 & v106 & 3p6p & 3d6d * & 4 & 6 & 5.54 + 4142.94 & 0.0550 & 0.0598 & & o ii & 4143.74 & v106 & 3p6p & 3d6d * & 6 & 8 & 0.73 + 4142.96 & 0.2610 & 0.2840 & & he i & 4143.76 & v53 & 2p 1p * & 6d 1d & 3 & 5 & 0.65 + 4143.44 & 0.0300 & 0.0326 & & [ fe iv ] & 4144.20 & & 3d5 4 g & 3d5 2h & 12 & 12 & 0.67 + 4145.02 & 0.0170 & 0.0185 & & n ii & 4145.78 & v65 & 3s 5p & 3p 5s * & 7 & 5 & 17.65 + & & & * & fe iii & 4145.79 & & 4 g 5p5f * & 4 g 6s5 g & 9 & 9 & + 4145.32 & 0.0080 & 0.0087 & & o ii & 4146.08 & v106 & 3p6p & 3d6d * & 8 & 10 & : + 4145.79 & 0.0060 & 0.0065 & & [ fe ii ] & 4146.65 & & a4f & a4 g & 10 & 8 & 48.17 + & & & * & ne ii & 4150.69 & v54a & 3d 4d & 4f 1@xmath402@xmath41 * & 2 & 4 & + 4151.72 & 0.0130 & 0.0141 & & c iii & 4152.51 & v21 & 3p 3d & 5f 3f * & 3 & 5 & 12.31 + 4152.51 & 0.1030 & 0.1119 & & o ii & 4153.30 & v19 & 3p 4p * & 3d 4p & 4 & 6 & 5.83 + 4155.70 & 0.0180 & 0.0196 & & c iii & 4156.50 & v21 & 3p 3d & 5f 3f * & 5 & 7 & 13.33 + & & & * & n ii & 4156.39 & v50b & 3d 3d * & 4f 2@xmath402@xmath41 & 3 & 5 & + 4155.74 & 0.0520 & 0.0565 & & o ii & 4156.53 & v19 & 3p 4p * & 3d 4p & 6 & 4 & 11.49 + 4155.97 & 0.0020 & 0.0022 & & c iii & 4156.74 & v21 & 3p 3d & 5f 3f * & 5 & 5 & 15.00 + 4156.19 & 0.0020 & 0.0022 & & n ii & 4157.01 & v50b & 3d 3d * & 4f 2@xmath402@xmath41 & 3 & 3 & : + & & & * & n ii & 4160.50 & v50b & 3d 3d * & 4f 2@xmath402@xmath41 & 5 & 5 & + & & & * & n ii & 4161.14 & v50b & 3d 3d * & 4f 2@xmath402@xmath41 & 5 & 3 & + 4162.03 & 0.0040 & 0.0043 & & c iii & 4162.88 & v21 & 3p 3d & 5f 3f * & 7 & 9 & 80.00 + 4162.51 & 0.0060 & 0.0065 & & c iii & 4163.25 & v21 & 3p 3d & 5f 3f * & 7 & 7 & 53.33 + & & & * & [ k v ] & 4163.32 & & 3p3 4s * & 3p3 2d * & 4 & 4 & + 4164.94 & 0.0010 & 0.0011 & & n ii & 4165.77 & v50b & 3d 3d * & 4f 2@xmath402@xmath41 & 7 & 5 & : + 4166.12 & 0.0100 & 0.0108 & & fe iii & 4166.88 & & 4 g 5p5h * & 4 g 5d5 g & 11 & 9 & 17.00 + 4168.50 & 0.0720 & 0.0781 & & o ii & 4169.22 & v19 & 3p 4p * & 3d 4p & 6 & 6 & 2.92 + & & & * & he i & 4168.97 & v52 & 2p 1p * & 6s 1s & 3 & 1 & + & & & * & n ii & 4169.38 & v50a & 3d 3d * & 4f 2@xmath403@xmath41 & 3 & 5 & + 4170.85 & 0.0212 & 0.0230 & & n ii & 4171.61 & v43b & 3d 1d * & 4f 1@xmath404@xmath41 & 5 & 7 & 10.44 + & & & * & fe i & 4171.69 & v941 & b1 g & x1f * & 9 & 7 & + 4172.87 & 0.0032 & 0.0035 & & n
ii & 4173.57 & v50a & 3d 3d * & 4f 2@xmath403@xmath41 & 5 & 5 & 50.00 + & & & * & fe ii & 4173.46 & v27 & b4p & z4d * & 6 & 6 & + 4173.55 & 0.0009 & 0.0010 & & n ii & 4174.38 & v50a & 3d 3d * & 4f 2@xmath403@xmath41 & 5 & 7 & : + & & & * & n ii & 4175.66 & v43a & 3d 1d * & 4f 1@xmath403@xmath41 & 5 & 5 & + 4175.33 & 0.0260 & 0.0281 & & n ii & 4176.16 & v43a & 3d 1d * & 4f 1@xmath403@xmath41 & 5 & 7 & 12.76 + & & & * & fe iii & 4176.09 & & b2 g 4s3 g & 2h 4p3h * & 11 & 13 & + 4177.40 & 0.0080 & 0.0087 & & fe iii ] & 4178.13 & & 6s 4p7p * & b2d 4s3d & 5 & 7 & 23.75 + 4178.03 & 0.0010 & 0.0011 & & n ii & 4178.85 & v50a & 3d 3d * & 4f 2@xmath403@xmath41 & 7 & 5 & : + 4178.84 & 0.0094 & 0.0102 & & n ii & 4179.67 & v50a & 3d 3d * & 4f 2@xmath403@xmath41 & 7 & 7 & 27.14 + 4179.79 & 0.0106 & 0.0115 & & [ fe v ] & 4180.60 & & 3d4 5d & 3d4 3p2 & 3 & 1 & 3.33 + 4180.35 & 0.0350 & 0.0379 & & n ii & 4181.10 & v49b & 3d 3d * & 4f 2@xmath405@xmath41 & 7 & 9 & 2.57 + & & & * & fe ii & 4181.20 & & y4d * & e4f & 6 & 6 & + 4184.67 & 0.0650 & 0.0703 & & o ii & 4185.45 & v36 & 3p 2f * & 3d 2 g & 6 & 8 & 4.31 + 4186.19 & 0.0820 & 0.0887 & & c iii & 4186.90 & v18 & 4f 1f * & 5
g 1 g & 7 & 9 & 3.29 + & & & * & o ii & 4189.58 & v36 & 3p 2f * & 3d 2 g & 8 & 8 & + 4189.01 & 0.0770 & 0.0833 & & o ii & 4189.79 & v36 & 3p 2f * & 3d 2 g & 8 & 10 & 1.82 + 4191.75 & 0.0006 & 0.0007 & & o ii & 4192.51 & v42 & 3p 2d * & 3d 2p & 6 & 4 & : + 4194.90 & 0.0430 & 0.0465 & & n iii & 4195.76 & v6 & 3s 2p * & 3p 2d & 2 & 4 & 10.70 + 4195.16 & 0.0040 & 0.0042 & & n ii & 4195.97 & v49a & 3d 3d * & 4f 2@xmath404@xmath41 & 5 & 7 & : + 4197.43 & 0.0240 & 0.0259 & & [ fe iv ] & 4198.24 & & 3d5 4 g & 3d5 2h & 12 & 10 & 9.17 + 4198.97 & 0.2530 & 0.2732 & & he ii & 4199.83 & 4.11 & 4f+ 2f * & 11g+ 2 g & 32 & & 2.69 + 4199.18 & 0.0070 & 0.0075 & & n ii & 4199.98 & v49a & 3d 3d * & 4f 2@xmath404@xmath41 & 7 & 9 & : + & & & * & ar ii & 4199.89 & v124 & 3d 2d & 4f 0@xmath403@xmath41 * & 4 & 6 & + 4199.24 & 0.0780 & 0.0842 & & n iii & 4200.10 & v6 & 3s 2p * & 3p 2d & 4 & 6 & 17.82 + & & & * & n ii & 4201.35 & v49a & 3d 3d * & 4f 2@xmath404@xmath41 & 7 & 7 & + 4205.57 & 0.0050 & 0.0054 & & n ii & 4206.51 & v74 & 3d 5d & 4f 5f * & 7 & 9 & 22.00 + & & & * & [ fe iv ] & 4206.60 & & 3d5 4 g & 3d5 2h & 10 & 10 & + 4207.01 & 0.0013 & 0.0014 & & n ii & 4207.50 & v74 & 3d 5d & 4f 5f * & 9 & 11 & 70.00 + & & & * & [ fe iv ] & 4208.08 & & 3d5 4 g & 3d5 2h & 8 & 10 & + 4208.58 & 0.0020 & 0.0022 & & ne ii & 4209.33 & & 3p '' 2p * & 5s 2d & 4 & 6 & 35.00 + 4211.47 & 0.0012 & 0.0013 & & si iv & 4212.41 & v5 & 5d 2d & 6f 2f * & 10 & 14 & : + 4214.99 & 0.0080 & 0.0086 & & n iii & 4215.77 & v6 & 3s 2p * & 3p 2d & 4 & 4 & 20.00 + 4216.39 & 0.0110 & 0.0118 & & ne ii & 4217.17 & v52b & 3d 4d & 4f 2@xmath403@xmath41 * & 8 & 8 & 13.64 + & & & * & ne ii & 4217.15 & v52b & 3d 4d & 4f 2@xmath403@xmath41 * & 8 & 6 & + 4216.41 & 0.0020 & 0.0022 & & s ii & 4217.18 & v44 & 4p 4d * & 4d 4f & 8 & 8 & 5.00 + & & & * & ne ii & 4219.37 & v52a & 3d 4d & 4f 2@xmath404@xmath41 * & 8 & 8 & 15.86 + 4218.97 & 0.0531 & 0.0571 & & ne ii & 4219.74 & v52a & 3d 4d & 4f 2@xmath404@xmath41 * & 8 & 10 & 20.90 + 4220.11 & 0.0159 & 0.0171 & & ne ii & 4220.89 & v52d & 3d 4d & 4f 2@xmath402@xmath41 * & 6 & 6 & 18.67 + 4221.99 & 0.0030 & 0.0032 & & o i & 4222.76 & v33 & 4p 3p & 3d 3p * & 3 & 3 & 10.00 + & & & * & o i & 4222.76 & v33 & 4p 3p & 3d 3p * & 5 & 3 & + 4222.05 & 0.0040 & 0.0043 & & o i & 4222.82 & v33 & 4p 3p & 3d 3p * & 1 & 3 & 12.50 + 4223.51 & 0.0070 & 0.0075 & & ne ii & 4224.47 & v52c & 3d 4d & 4f 2@xmath401@xmath41 * & 4 & 4 & 2.86 + 4223.68 & 0.0020 & 0.0022 & & ne ii & 4224.64 & v52c & 3d 4d & 4f 2@xmath401@xmath41 * & 4 & 2 & : + 4225.96 & 0.0150 & 0.0161 & & [ fe v ] & 4227.19 & & 3d4 5d & 3d4 3h & 9 & 9 & 2.67 + 4226.71 & 0.2560 & 0.2754 & & n ii & 4227.69 & v33 & 3p 1d & 4s 1p * & 5 & 3 & 2.50 + & & & * & ne ii & 4231.53 & v52b & 3d 4d & 4f 2@xmath403@xmath41 * & 6 & 6 & + & & & * & [ fe ii ] & 4231.55 & & 3d7 4f & ( 3g)4s 4 g & 8 & 6 & + 4230.68 & 0.0256 & 0.0275 & & ne ii & 4231.64 & v52b & 3d 4d & 4f 2@xmath403@xmath41 * & 6 & 8 & 15.59 + 4232.88 & 0.0164 & 0.0175 & & ne ii & 4233.85 & v52a & 3d 4d & 4f 2@xmath404@xmath41 * & 6 & 8 & 16.00 + 4236.08 & 0.0510 & 0.0548 & & n ii & 4236.91 & v48a & 3d 3d * & 4f 1@xmath403@xmath41 & 3 & 5 & 11.11 + & & & * & n ii & 4237.05 & v48b & 3d 3d * & 4f 1@xmath404@xmath41 & 5 & 7 & + 4237.10 & 0.0110 & 0.0118 & & c iii & 4237.91 & & ? & ? & & & 57.66 + 4239.09 & 0.0160 & 0.0172 & & ne ii & 4239.91 & v52d & 3d 4d & 4f 2@xmath402@xmath41 * & 4 & 6 & 11.25 + & & & * & ne ii & 4240.10 & v52d & 3d 4d & 4f 2@xmath402@xmath41 * & 4 & 4 & + 4240.41 & 0.0001 & 0.0002 & & n ii & 4241.24 & v48a & 3d 3d * & 4f 1@xmath403@xmath41 & 5 & 5 & : + 4240.95 & 0.0302 & 0.0325 & & n ii & 4241.78 & v48a & 3d 3d * & 4f 1@xmath403@xmath41 & 5 & 7 & 36.09 + 4240.95 & 0.0565 & 0.0606 & & n ii & 4241.78 & v48b & 3d 3d * & 4f 1@xmath404@xmath41 & 7 & 9 & 4.67 + & & & * & ne ii & 4242.04 & v52c & 3d 4d & 4f 2@xmath401@xmath41 * & 2 & 4 & + 4241.66 & 0.0020 & 0.0021 & & n ii & 4242.49 & v48b & 3d 3d * & 4f 1@xmath404@xmath41 & 7 & 7 & 5.00 + 4243.31 & 0.0060 & 0.0064 & & ne ii & 4244.14 & v62 & 3d 2d & 4f 0@xmath403@xmath41 * & 6 & 8 & 28.33 + & & & * & ne ii & 4244.17 & v62 & 3d 2d & 4f 0@xmath403@xmath41 * & 6 & 6 & + & & & * & [ fe ii ] & 4243.97 & & 3d7 4f & ( 3g)4s 4 g & 10 & 12 & + 4244.74 & 0.0030 & 0.0032 & & [ fe ii ] & 4244.81 & & 3d7 4f & ( 3g)4s 4 g & 8 & 8 & 83.33 + 4245.76 & 0.0008 & 0.0009 & & n ii & 4246.71 & v48a & 3d 3d * & 4f 1@xmath403@xmath41 & 7 & 5 & : + 4246.25 & 0.0040 & 0.0043 & & n ii & 4247.20 & v48a & 3d 3d * & 4f 1@xmath403@xmath41 & 7 & 7 & 12.50 + 4249.61 & 0.0150 & 0.0161 & & ne ii & 4250.65 & v52b & 3d 4d & 4f 2@xmath403@xmath41 * & 4 & 6 & 51.33 + 4250.99 & 0.0090 & 0.0096 & & [ fe ii ] & 4251.44 & & 3d7 4f & ( 3h)4s 2h & 6 & 10 & 16.67 + 4253.12 & 0.0542 & 0.0581 & & o ii & 4253.90 & v101 & 3d 2 g & 4f h5 * & 10 & 12 & 16.90 + & & & * & o ii & 4253.91 & v101 & 3d 2 g & 4f h5 * & 10 & 10 & + & & & * & o ii & 4254.12 & v101 & 3d 2 g & 4f h5 * & 8 & 10 & + 4254.55 & 0.0040 & 0.0043 & & c iii & 4255.41 & v15.01 & 4f 3f * & 5d 3d & 5 & 3 & : + 4255.59 & 0.0040 & 0.0043 & & c iii & 4256.46 & v15.01 & 4f 3f * & 5d 3d & 7 & 5 & : + 4257.00 & 0.0150 & 0.0161 & & ne ii & 4257.80 & v52d & 3d 4d & 4f 2@xmath402@xmath41 * & 2 & 4 & 79.33 + 4259.02 & 0.0050 & 0.0054 & & ni ii & 4259.99 & & 4fsp4 g * & 3f 9s4f & 6 & 6 & : + 4262.54 & 0.0040 & 0.0043 & & o ii & 4263.27 & v53c & 3d 4p & 4f d1 * & 6 & 4 & 2.50 + 4266.27 & 0.3290 & 0.3520 & & c ii & 4267.00 & v6 & 3d 2d & 4f 2f * & 4 & 6 & 1.85 + 4266.41 & 0.4700 & 0.5029 & & c ii & 4267.26 & v6 & 3d 2d & 4f 2f * & 6 & 8 & 1.85 + 4266.53 & 0.0230 & 0.0246 & & c ii & 4267.26 & v6 & 3d 2d & 4f 2f * & 6 & 6 & 1.74 + 4266.65 & 0.0040 & 0.0043 & & ne ii & 4267.38 & v57c & 3d 4f & 4f 1@xmath403@xmath41 * & 10 & 8 & : + 4272.26 & 0.0110 & 0.0118 & & o ii & 4273.10 & v67a & 3d 4d & 4f f4 * & 6 & 8 & 5.45 + 4273.40 & 0.0180 & 0.0192 & & o ii & 4274.24 & v67a & 3d 4d & 4f f4 * & 8 & 8 & 28.89 + 4274.71 & 0.1030 & 0.1101 & & o ii & 4275.55 & v67a & 3d 4d & 4f f4 * & 8 & 10 & 4.95 + 4275.15 & 0.0230 & 0.0246 & & o ii & 4275.99 & v67b & 3d 4d & 4f f3 * & 4 & 6 & 4.78 + 4275.44 & 0.0160 & 0.0171 & & o ii & 4276.28 & v67b & 3d 4d & 4f f3 * & 6 & 6 & 5.00 + 4275.78 & 0.0150 & 0.0160 & & o ii & 4276.62 & v53c & 3d 4p & 4f d1 * & 4 & 4 & 5.33 + & & & * & o ii & 4276.71 & v53c & 3d 4p & 4f d1 * & 4 & 2 & + 4275.91 & 0.0500 & 0.0534 & & o ii & 4276.75 & v67b & 3d 4p & 4f d1 * & 6 & 8 & 4.80 + 4276.59 & 0.0320 & 0.0342 & & o ii & 4277.43 & v67c & 3d 4d & 4f f2 * & 2 & 4 & 5.00 + & & & * & o ii & 4277.44 & v67b & 3d 4d & 4f f3 * & 8 & 6 & + 4277.05 & 0.0210 & 0.0224 & & o ii & 4277.89 & v67b & 3d 4d & 4f f3 * & 8 & 8 & 4.76 + & & & * & ne ii & 4280.44 & v57b & 3d 4f & 4f 1<4 > * & 10 & 8 & + & & & * & ne ii & 4280.59 & v57b & 3d 4f & 4f 1<4 > * & 10 & 10 & + 4280.61 & 0.0200 & 0.0214 & & o ii & 4281.32 & v53b & 3d 4p & 4f d2 * & 6 & 6 & 2.50 + 4280.75 & 0.0014 & 0.0015 & & o ii & 4281.46 & v53b & 3d 4p & 4f d2 * & 6 & 4 & : + & & & * & o ii & 4282.02 & v78a & 3d 2f & 4f f4 * & 6 & 8 & + 4282.25 & 0.0390 & 0.0416 & & o ii & 4282.96 & v67c & 3d 4d & 4f f2 * & 4 & 6 & 2.31 + 4282.54 & 0.0070 & 0.0075 & & o ii & 4283.25 & v67c & 3d 4d & 4f f2 * & 6 & 6 & 2.86 + 4283.02 & 0.0210 & 0.0224 & & ne ii & 4283.73 & v57c & 3d 4f & 4f 1@xmath403@xmath41 * & 8 & 8 & 2.38 + & & & * & o ii & 4283.73 & v67c & 3d 4d & 4f f2 * & 4 & 4 & + 4283.29 & 0.0020 & 0.0021 & & o ii & 4284.00 & v67c & 3d 4d & 4f f2 * & 6 & 4 & : + 4283.68 & 0.0013 & 0.0014 & & o ii & 4284.39 & v67c & 3d 4d & 4f f2 * & 8 & 6 & : + & & & * & [ ni ii ] & 4285.31 & & 3d9 2d & ( 3p)4s 4p & 4 & 2 & + & & & * & o ii & 4285.21 & v78b & 3d 2f & 4f f3 * & 6 & 6 & + 4284.98 & 0.0570 & 0.0609 & & o ii & 4285.69 & v78b & 3d 2f & 4f f3 * & 6 & 8 & 7.37 + 4288.09 & 0.0170 & 0.0181 & & o ii & 4288.82 & v53c & 3d 4p & 4f d1 * & 2 & 4 & 7.65 + 4288.09 & 0.0190 & 0.0203 & & o ii & 4288.82 & v53c & 3d 4p & 4f d1 * & 2 & 2 & 7.89 + 4290.52 & 0.0550 & 0.0587 & & o ii & 4291.25 & v55 & 3d 4p & 4f g3 * & 6 & 8 & 8.54 + 4291.13 & 0.0030 & 0.0032 & & o ii & 4291.86 & v55 & 3d 4p & 4f g3 * & 6 & 6 & : + 4291.48 & 0.0220 & 0.0235 & & o ii & 4292.21 & v78c & 3d 2f & 4f f2 * & 6 & 6 & 3.18 + 4292.22 & 0.0080 & 0.0085 & & o ii & 4292.95 & v78c & 3d 2f & 4f f2 * & 6 & 4 & 42.50 + 4294.05 & 0.0630 & 0.0672 & & o ii & 4294.78 & v53b & 3d 4p & 4f d2 * & 4 & 6 & 3.17 + 4294.19 & 0.0180 & 0.0192 & & o ii & 4294.92 & v53b & 3d 4p & 4f d2 * & 4 & 4 & 3.33 + 4296.14 & 0.0030 & 0.0032 & & ne ii & 4296.87 & v57b & 3d 4f & 4f 1<4 > * & 8 & 8 & : + & & & * & ne ii & 4297.02 & v57b & 3d 4f & 4f 1<4 > * & 8 & 10 & + 4302.79 & 0.0200 & 0.0213 & & o ii & 4303.61 & v65a & 3d 4d & 4f g5 * & 8 & 10 & 3.00 + 4303.01 & 0.1130 & 0.1203 & & o ii & 4303.82 & v53a & 3d 4p & 4f d3 * & 6 & 8 & 3.27 + 4303.26 & 0.0050 & 0.0053 & & o ii & 4304.08 & v53a & 3d 4p & 4f d3 * & 6 & 6 & 4.00 + 4304.57 & 0.0060 & 0.0064 & & o ii & 4305.39 & v55 & 3d 4p & 4f g3 * & 4 & 6 & 3.33 + 4306.41 & 0.0270 & 0.0287 & & o ii & 4307.23 & v53b & 3d 4p & 4f d2 * & 2 & 4 & 10.00 + 4308.17 & 0.0150 & 0.0160 & & o ii & 4309.00 & v63c & 3d 4d & 4f d1 * & 2 & 4 & 18.00 + & & & * & o ii & 4309.00 & v63c & 3d 4d & 4f d1 * & 2 & 2 & + 4311.35 & 0.0150 & 0.0160 & & o ii & 4312.11 & v78a & 3d 2f & 4f f4 * & 8 & 8 & 16.67 + 4312.68 & 0.0230 & 0.0245 & & o ii & 4313.44 & v78a & 3d 2f & 4f f4 * & 8 & 10 & 10.87 + 4314.63 & 0.0040 & 0.0043 & & o ii & 4315.39 & v63c & 3d 4d & 4f d1 * & 4 & 2 & 2.50 + 4314.63 & 0.0120 & 0.0128 & & o ii & 4315.39 & v78b & 3d 2f & 4f f3 * & 8 & 6 & 33.33 + 4314.64 & 0.0060 & 0.0064 & & o ii & 4315.40 & v63c & 3d 4d & 4f d1 * & 4 & 4 & 16.67 + 4314.93 & 0.0012 & 0.0013 & & o ii & 4315.69 & v63c & 3d 4d & 4f d1 * & 6 & 4 & : + 4315.07 & 0.0090 & 0.0096 & & o ii & 4315.83 & v78b & 3d 2f & 4f f3 * & 8 & 8 & 2.22 + 4316.38 & 0.0860 & 0.0914 & & o ii & 4317.14 & v2 & 3s 4p & 3p 4p * & 2 & 4 & 2.21 + 4316.94 & 0.0150 & 0.0159 & & o ii & 4317.70 & v53a & 3d 4p & 4f d3 * & 4 & 6 & 2.00 + 4318.87 & 0.0590 & 0.0627 & & o ii & 4319.63 & v2 & 3s 4p & 3p 4p * & 4 & 6 & 4.58 + 4321.69 & 0.0010 & 0.0011 & & o ii & 4322.45 & v78c & 3d 2f & 4f f2 * & 8 & 6 & : + 4324.92 & 0.0270 & 0.0287 & & o ii & 4325.76 & v2 & 3s 4p & 3p 4p * & 2 & 2 & 6.67 + 4326.75 & 0.0050 & 0.0053 & & o ii & 4327.59 & v63b & 3d 4d & 4f d2 * & 2 & 4 & 30.00 + 4328.62 & 0.0130 & 0.0138 & & [ fe ii ] & 4329.43 & & ( 5d)4s 4d & ( 3d)4s 4d & 6 & 8 & : + 4330.32 & 0.0220 & 0.0233 & & o ii & 4331.13 & v65b & 3d 4d & 4f g4 * & 6 & 8 & 1.36 + & & & * & o ii & 4332.33 & v65b & 3d 4d & 4f g4 * & 8 & 8 & + 4332.02 & 0.0550 & 0.0584 & & o ii & 4332.71 & v65b & 3d 4d & 4f g4 * & 8 & 10 & 1.27 + 4333.07 & 0.0070 & 0.0074 & & o ii & 4333.88 & v63b & 3d 4d & 4f d2 * & 4 & 6 & 1.43 + 4333.38 & 0.0080 & 0.0085 & & o ii & 4334.19 & v63b & 3d 4d & 4f d2 * & 6 & 6 & 1.25 + & & & * & o ii & 4334.03 & v63b & 3d 4d & 4f d2 * & 4 & 4 & + 4333.52 & 0.0020 & 0.0021 & & o ii & 4334.33 & v63b & 3d 4d & 4f d2 * & 6 & 4 & : + 4334.55 & 0.0020 & 0.0021 & & o ii & 4335.36 & v63b & 3d 4d & 4f d2 * & 8 & 6 & : + 4336.05 & 0.0500 & 0.0530 & & o ii & 4336.86 & v2 & 3s 4p & 3p 4p * & 4 & 4 & 1.20 + 4337.86 & 0.3410 & 0.3615 & & he ii & 4338.67 & 4.10 & 4f+ 2f * & 10g+ 2 g & 32 & & 1.29 + 4339.52 & 0.0800 & 0.0848 & & o ii & 4340.33 & v76b & 3d 2f & 4f g4 * & 6 & 8 & 1.25 + 4339.62 & 45.5200 & 48.2491 & & h i & 4340.46 & h5 & 2p+ 2p * & 5d+ 2d & 8 & 50 & 1.14 + 4341.19 & 0.1980 & 0.2098 & & o ii & 4342.00 & v76a & 3d 2f & 4f g5 * & 8 & 10 & 1.26 + 4343.61 & 0.0290 & 0.0307 & & o ii & 4344.42 & v65c & 3d 4d & 4f g3 * & 6 & 8 & 1.38 + & & & * & o ii & 4344.68 & v65c & 3d 4d & 4f g3 * & 4 & 6 & + 4344.15 & 0.0008 & 0.0009 & & o ii & 4344.96 & v65c & 3d 4d & 4f g3 * & 6 & 6 & : + 4344.62 & 0.0320 & 0.0339 & & o ii & 4345.55 & v65c & 3d 4d & 4f g3 * & 8 & 8 & 2.19 + 4344.74 & 0.1330 & 0.1409 & & o ii & 4345.56 & v2 & 3s 4p & 3p 4p * & 4 & 2 & 1.43 + 4345.34 & 0.0004 & 0.0005 & & o ii & 4346.16 & v65c & 3d 4d & 4f g3 * & 8 & 6 & : + 4346.03 & 0.0490 & 0.0519 & & [ fe ii ] & 4346.85 & & 3d7 4f & ( 3g)4s 4 g & 8 & 12 & 14.90 + 4346.40 & 0.0005 & 0.0006 & & o ii & 4347.22 & v16 & 3s 2d & 3p 2d * & 6 & 4 & : + 4346.59 & 0.0050 & 0.0053 & & o ii & 4347.42 & v16 & 3s 2d & 3p 2d * & 4 & 4 & 2.00 + 4348.60 & 0.1840 & 0.1948 & & o ii & 4349.43 & v2 & 3s 4p & 3p 4p * & 6 & 6 & 2.88 + 4350.44 & 0.0110 & 0.0116 & & o ii & 4351.26 & v16 & 3s 2d & 3p 2d * & 6 & 6 & 2.73 + 4350.63 & 0.0001 & 0.0002 & & o ii & 4351.45 & v16 & 3s 2d & 3p 2d * & 4 & 6 & : + 4350.99 & 0.0130 & 0.0138 & & [ fe ii ] & 4351.81 & & ( 5d)4s 4d & ( 3d)4s 4d & 6 & 4 & 3.08 + 4351.96 & 0.0020 & 0.0021 & & [ fe ii ] & 4352.78 & & 3d7 4f & ( 3g)4s 4 g & 6 & 10 & : + 4352.97 & 0.0240 & 0.0254 & & o ii & 4353.59 & v76c & 3d 2f & 4f g3 * & 6 & 8 & 21.25 + 4353.36 & 0.0005 & 0.0006 & & o ii & 4354.18 & v76c & 3d 2f & 4f g3 * & 6 & 6 & : + & & & * & o ii & 4357.25 & v18 & 3p 4p * & 3d 4f & 6 & 8 & + 4356.41 & 0.0235 & 0.0249 & & o ii & 4357.25 & v63a & 3d 4d & 4f d3 * & 6 & 8 & 5.24 + & & & * & o ii & 4357.52 & v63a & 3d 4d & 4f d3 * & 6 & 6 & + & & & * & [ fe ii ] & 4358.36 & & 3d7 4f & ( 3g)4s 4 g & 4 & 6 & + 4357.61 & 0.0020 & 0.0021 & & o ii & 4358.45 & v63a & 3d 4d & 4f d3 * & 8 & 8 & 5.00 + 4358.49 & 0.0280 & 0.0296 & & [ fe ii ] & 4359.33 & & ( 5d)4s 6d & 4s2 6s & 8 & 6 & 5.36 + & & & * & o ii & 4359.39 & v26 & 3p 2d * & 3d 2d & 4 & 6 & + 4362.37 & 6.9410 & 7.3361 & & [ o iii ] & 4363.21 & 2f & 2p2 1d & 2p2 1s & 5 & 1 & 5.82 + 4364.24 & 0.0020 & 0.0021 & & o iii & 4365.08 & v39b & 4f f@xmath403@xmath41 & 5 g g@xmath405@xmath41 * & 7 & 9 & 5.00 + & & & * & ne ii & 4365.64 & v60c & 3d 2f & 4f 1@xmath403@xmath41 * & 8 & 6 & + & & & * & ne ii & 4365.75 & v60c & 3d 2f & 4f 1@xmath403@xmath41 * & 8 & 8 & + 4365.54 & 0.0001 & 0.0002 & & ni ii & 4366.52 & & 3f 4d4p & 3p 5p4s * & 6 & 4 & : + 4365.69 & 0.0370 & 0.0391 & & o ii & 4366.53 & v75a & 3d 2f & 4f d3 * & 6 & 8 & 5.41 + & & & * & o ii & 4366.79 & v75a & 3d 2f & 4f d3 * & 6 & 6 & + 4366.15 & 0.0810 & 0.0856 & & o ii & 4366.89 & v2 & 3s 4p & 3p 4p * & 6 & 4 & 11.48 + & & & * & o iii & 4366.99 & v39a & 4f f@xmath403@xmath41 & 5 g g@xmath404@xmath41 * & 7 & 16 & + 4367.42 & 0.0140 & 0.0148 & & n i & 4368.40 & & 3p 2s * & 13s 2p & 2 & 4 & 5.00 + & & & * & c ii & 4368.26 & v45 & 3d 4p * & 4f 4d & 6 & 4 & + 4368.44 & 0.0030 & 0.0032 & & o ii & 4369.28 & v26 & 3p 2d * & 3d 2d & 4 & 4 & 36.67 + 4369.02 & 0.0200 & 0.0211 & & o iii & 4369.64 & v39a & 4f f@xmath403@xmath41 & 5 g g@xmath404@xmath41 * & 5 & 7 & 5.50 + 4370.81 & 0.0318 & 0.0336 & & o ii & 4371.62 & v76b & 3d 2f & 4f g4 * & 8 & 10 & 41.54 + 4373.47 & 0.0090 & 0.0095 & & o ii & 4374.30 & v75b & 3d 2f & 4f d2 * & 8 & 6 & 66.67 + & & & * & c ii & 4374.27 & v45 & 3d 4p * & 4f 4d & 6 & 8 & + 4375.76 & 0.0310 & 0.0327 & & o iii & 4376.54 & v39c & 4f f@xmath404@xmath41 & 5 g g@xmath405@xmath41 * & 7 & 9 & 25.81 + & & & * & c ii & 4376.56 & v45 & 3d 4p * & 4f 4d & 4 & 6 & + 4377.18 & 0.0360 & 0.0380 & & ne ii & 4377.98 & v65 & 3d 4p & 4f 0@xmath403@xmath41 * & 4 & 6 & 55.00 + 4378.31 & 0.3484 & 0.3675 & & n iii & 4379.11 & v18 & 4f 2f * & 5
g 2 g & 16 & 18 & 6.32 + & & & * & o iii & 4379.58 & v39c & 4f f@xmath404@xmath41 & 5 g g@xmath405@xmath41 * & 9 & 20 & + 4378.74 & 0.0470 & 0.0496 & & ne ii & 4379.55 & v60b & 3d 2f & 4f 1@xmath404@xmath41 * & 8 & 10 & 65.32 + & & & * & ne ii & 4379.40 & v60b & 3d 2f & 4f 1@xmath404@xmath41 * & 8 & 8 & + 4384.22 & 0.0300 & 0.0316 & & ne ii & 4385.06 & v59 & 3d 2f & 4f 0@xmath403@xmath41 * & 6 & 8 & 18.33 + & & & * & o ii & 4384.70 & v76c & 3d 2f & 4f g3 * & 8 & 8 & + & & & * & ne ii & 4384.98 & v59 & 3d 2f & 4f 0@xmath403@xmath41 * & 6 & 6 & + & & & * & o ii & 4385.33 & v76c & 3d 2f & 4f g3 * & 8 & 6 & + 4387.09 & 0.5680 & 0.5985 & & he i & 4387.93 & v51 & 2p 1p * & 5d 1d & 3 & 5 & 1.74 + 4391.16 & 0.0707 & 0.0744 & & ne ii & 4391.99 & v55e & 3d 4f & 4f 2@xmath405@xmath41 * & 10 & 12 & 5.10 + & & & * & ne ii & 4392.00 & v55e & 3d 4f & 4f 2@xmath405@xmath41 * & 10 & 10 & + 4395.10 & 0.0070 & 0.0074 & & o ii & 4395.94 & v26 & 3p 2d * & 3d 2d & 6 & 6 & 17.50 + 4397.15 & 0.0230 & 0.0242 & & ne ii & 4397.99 & v57b & 3d 4f & 4f 1@xmath404@xmath41 * & 6 & 8 & 11.25 + 4397.87 & 0.0020 & 0.0021 & & ni ii & 4398.42 & & 4f sp4d * & 3f 5d2f & 6 & 8 & : + 4399.91 & 0.0040 & 0.0042 & & ne ii & 4401.67 & v57a & 3d 4f & 4f 1@xmath402@xmath41 * & 6 & 6 & : + & & & * & ne ii & 4401.88 & v57a & 3d 4f & 4f 1@xmath402@xmath41 * & 6 & 4 & + 4405.14 & 0.0001 & 0.0001 & & o ii & 4405.99 & v26 & 3p 2d * & 3d 2d & 6 & 4 & : + 4407.44 & 0.0070 & 0.0074 & & o iii & 4408.29 & v46a & 4f g@xmath404@xmath41 & 5 g h@xmath405@xmath41 * & 7 & 9 & 31.43 + 4408.50 & 0.0569 & 0.0597 & & ne ii & 4409.30 & v55e & 3d 4f & 4f 2@xmath405@xmath41 * & 8 & 10 & 8.68 + & & & * & ne ii & 4409.78 & v55b & 3d 4f & 4f 2@xmath403@xmath41 * & 10 & 8 & + 4411.40 & 0.0040 & 0.0042 & & ne ii & 4412.20 & v55a & 3d 4f & 4f 2@xmath404@xmath41 * & 10 & 8 & : + 4411.79 & 0.0130 & 0.0137 & & ne ii & 4412.59 & v55a & 3d 4f & 4f 2@xmath404@xmath41 * & 10 & 10 & 36.15 + 4412.42 & 0.0267 & 0.0280 & & ne ii & 4413.22 & v65 & 3d 4p & 4f 0@xmath403@xmath41 * & 6 & 8 & 2.00 + & & & * & ne ii & 4413.11 & v65 & 3d 4p & 4f 0@xmath403@xmath41 * & 6 & 6 & + & & & * & ne ii & 4413.11 & v57c & 3d 4f & 4f 1@xmath403@xmath41 * & 4 & 6 & + 4414.00 & 0.0080 & 0.0084 & & o iii & 4414.85 & v46a & 4f g@xmath404@xmath41 & 5 g h@xmath405@xmath41 * & 9 & 11 & 32.50 + 4414.05 & 0.0950 & 0.0998 & & o ii & 4414.90 & v5 & 3s 2p & 3p 2d * & 4 & 6 & 1.89 + 4415.96 & 0.0240 & 0.0252 & & ne ii & 4416.76 & v61d & 3d 2d & 4f 2@xmath402@xmath41 * & 6 & 6 & 10.83 + & & & * & ne ii & 4416.77 & v61d & 3d 2d & 4f 2@xmath402@xmath41 * & 6 & 4 & + 4416.12 & 0.0610 & 0.0640 & & o ii & 4416.97 & v5 & 3s 2p & 3p 2d * & 2 & 4 & 0.98 + 4416.22 & 0.0040 & 0.0042 & & n ii & 4417.07 & v55b & 3d 3p * & 4f 2@xmath402@xmath41 & 5 & 5 & : + 4416.97 & 0.0020 & 0.0021 & & n ii & 4417.83 & v55b & 3d 3p * & 4f 2@xmath402@xmath41 & 5 & 3 & : + 4420.51 & 0.0080 & 0.0084 & & ne ii & 4421.39 & v66c & 3d 4p & 4f 1@xmath403@xmath41 * & 4 & 6 & 2.50 + 4426.40 & 0.0090 & 0.0094 & & n ii & 4427.24 & v55b & 3d 3p * & 4f 2@xmath402@xmath41 & 3 & 5 & 18.89 + & & & * & ne
ii & 4427.22 & v55b & 3d 4f & 4f 2@xmath403@xmath41 * & 8 & 8 & + 4427.12 & 0.0060 & 0.0063 & & n ii & 4427.96 & v55b & 3d 3p * & 4f 2@xmath402@xmath41 & 3 & 3 & 20.00 + & & & * & ne ii & 4428.52 & v61b & 3d 2d & 4f 2@xmath403@xmath41 * & 6 & 8 & + & & & * & ne ii & 4428.41 & v61b & 3d 2d & 4f 2@xmath403@xmath41 * & 6 & 6 & + & & & * & ne ii & 4428.52 & v60c & 3d 2f & 4f 1@xmath403@xmath41 * & 6 & 6 & + 4427.80 & 0.0420 & 0.0440 & & ne ii & 4428.64 & v60c & 3d 2f & 4f 1@xmath403@xmath41 * & 6 & 8 & 6.95 + & & & * & ne ii & 4430.06 & v55a & 3d 4f & 4f 2@xmath404@xmath41 * & 8 & 10 & + & & & * & ne ii & 4430.90 & v57a & 3d 4f & 4f 1@xmath402@xmath41 * & 4 & 6 & + 4430.20 & 0.0380 & 0.0399 & & ne ii & 4430.94 & v61a & 3d 2d & 4f 2@xmath404@xmath41 * & 6 & 8 & : + & & & * & ne ii & 4431.11 & v57a & 3d 4f & 4f 1@xmath402@xmath41 * & 4 & 4 & + 4430.98 & 0.0030 & 0.0031 & & n ii & 4431.82 & v55a & 3d 3p * & 4f 2@xmath403@xmath41 & 5 & 5 & 20.00 + 4431.48 & 0.0020 & 0.0021 & & [ fe ii ] & 4432.45 & & ( 5d)4s 6d & ( 3f2)4s 4f & 8 & 6 & : + 4431.90 & 0.0340 & 0.0356 & & n ii & 4432.74 & v55a & 3d 3p * & 4f 2@xmath403@xmath41 & 5 & 7 & 18.24 + 4432.64 & 0.0080 & 0.0084 & & n ii & 4433.48 & v55b & 3d 3p * & 4f 2@xmath402@xmath41 & 1 & 3 & 18.75 + 4433.76 & 0.0230 & 0.0241 & & o iii & 4434.60 & v46b & 4f g@xmath405@xmath41 & 5 g h@xmath406@xmath41 * & 11 & 24 & 10.00 + 4436.65 & 0.0010 & 0.0010 & & o iii & 4437.54 & v49b & 4f d@xmath403@xmath41 & 5 g f@xmath403@xmath41 * & 7 & 12 & 10.00 + 4436.66 & 0.0620 & 0.0649 & & he i & 4437.55 & v50 & 2p 1p * & 5s 1s & 3 & 1 & 71.94 + 4438.36 & 0.0130 & 0.0136 & & ne ii & 4439.25 & v66a & 3d 4p & 4f 1@xmath402@xmath41 * & 4 & 6 & 18.46 + & & & * & ne ii & 4439.46 & v66a & 3d 4p & 4f 1@xmath402@xmath41 * & 4 & 4 & + 4441.26 & 0.0190 & 0.0198 & & n ii & 4442.02 & v55a & 3d 3p * & 4f 2@xmath403@xmath41 & 3 & 5 & 9.17 + & & & * & o iii & 4442.02 & v49b & 4f d@xmath403@xmath41 & 5 g f@xmath403@xmath41 * & 5 & 12 & + & & & * & ne ii & 4442.69 & v60b & 3d 2f & 4f 1@xmath404@xmath41 * & 6 & 8 & + 4446.05 & 0.0050 & 0.0052 & & o ii & 4446.81 & v86c & 3d 2p & 4f d1 * & 4 & 4 & 40.00 + & & & * & n ii & 4447.03 & v15 & 3p 1p & 3d 1d * & 3 & 5 & + 4447.38 & 0.0370 & 0.0387 & & o iii & 4448.14 & v46b & 4f g@xmath405@xmath41 & 5 g h@xmath406@xmath41 * & 9 & 11 & 19.19 + 4447.43 & 0.0020 & 0.0021 & & o ii & 4448.19 & v35 & 3p 2f * & 3d 2f & 8 & 8 & 30.00 + 4451.30 & 0.0135 & 0.0141 & & [ fe ii ] & 4452.10 & & 5d 4s a6d & 3d5 4s2 a6 & 4 & 6 & 18.15 + 4451.31 & 0.0135 & 0.0141 & & o ii & 4452.37 & v5 & 3s 2p & 3p 2d * & 4 & 4 & 18.15 + 4453.27 & 0.0110 & 0.0115 & & o iii & 4454.03 & v49a & 4f d@xmath403@xmath41 & 5 g f@xmath404@xmath41 * & 7 & 16 & 42.73 + 4456.25 & 0.0250 & 0.0261 & & ne ii & 4457.05 & v61d & 3d 2d & 4f 2@xmath402@xmath41 * & 4 & 6 & 9.20 + & & & * & ne
ii & 4457.24 & v66c & 3d 4p & 4f 1@xmath403@xmath41 * & 6 & 6 & + & & & * & ne ii & 4457.24 & v61d & 3d 2d & 4f 2@xmath402@xmath41 * & 4 & 4 & + & & & * & ne ii & 4457.36 & v66c & 3d 4p & 4f 1@xmath403@xmath41 * & 6 & 8 & + & & & * & [ fe ii ] & 4457.95 & & 3d6 4s a6d & 3d6 4s b4f & 8 & 8 & + 4457.75 & 0.0170 & 0.0178 & & o iii & 4458.55 & v49a & 4f d@xmath403@xmath41 & 5 g f@xmath404@xmath41 * & 5 & 7 & 34.12 + 4460.32 & 0.0120 & 0.0125 & & c iii & 4461.14 & & ? & ? & & & 28.70 + 4465.60 & 0.0310 & 0.0324 & & o ii & 4466.42 & v86b & 3d 2p & 4f d2 * & 4 & 6 & 33.23 + 4465.77 & 0.0310 & 0.0324 & & o ii & 4466.59 & v86b & 3d 2p & 4f d2 * & 4 & 4 & 30.65 + 4468.64 & 0.2950 & 0.3078 & & o ii & 4469.46 & v86c & 3d 2p & 4f d1 * & 2 & 2 & : + & & & * & ne ii & 4468.91 & v61b & 3d 2d & 4f 2@xmath403@xmath41 * & 4 & 6 & + 4468.66 & 0.0360 & 0.0376 & & o ii & 4469.48 & v86c & 3d 2p & 4f d1 * & 2 & 4 & : + 4470.20 & 0.0050 & 0.0052 & & o iii & 4471.02 & v49c & 4f d@xmath402@xmath41 & 5 g f@xmath403@xmath41 * & 3 & 5 & : + & & & * & ne ii & 4471.04 & & 4d 4d & 8f 1@xmath404@xmath41 * & 8 & 10 & + 4470.66 & 4.5800 & 4.7771 & & he i & 4471.50 & v14 & 2p 3p * & 4d 3d & 9 & 15 & 9.71 + 4475.29 & 0.0040 & 0.0042 & & o iii & 4476.11 & v45b & 4f g@xmath404@xmath41 & 5 g g@xmath405@xmath41 * & 7 & 9 & : + & & & * & o iii & 4475.17 & v49c & 4f d@xmath402@xmath41 & 5 g f@xmath403@xmath41 * & 5 & 12 & + 4477.08 & 0.0289 & 0.0301 & & o ii & 4477.90 & v88 & 3d 2p & 4f g3 * & 4 & 6 & 54.55 + & & & * & ne ii & 4477.93 & & 3d 2d & 4f2@xmath401@xmath41 * & 6 & 4 & + & & & * & o iii & 4477.91 & v45a & 4f g@xmath404@xmath41 & 5 g g@xmath404@xmath41 * & 7 & 9 & + 4480.38 & 0.0290 & 0.0302 & & mg ii & 4481.20 & v4 & 3d 2d & 4f 2f * & 10 & 14 & 57.93 + 4482.01 & 0.0110 & 0.0115 & & o iii & 4482.83 & v45b & 4f g@xmath404@xmath41 & 5 g g@xmath405@xmath41 * & 9 & 20 & 52.73 + 4486.91 & 0.0130 & 0.0135 & & o ii & 4487.72 & v104 & 3d 2p & 4f d2 * & 2 & 4 & : + 4487.39 & 0.0140 & 0.0146 & & o ii & 4488.20 & v104 & 3d 2p & 4f d2 * & 4 & 6 & : + & & & * & o ii & 4488.20 & v104 & 3d 2p & 4f d2 * & 4 & 4 & + 4488.68 & 0.0210 & 0.0219 & & o ii & 4489.49 & v86b & 3d 2p & 4f d2 * & 2 & 4 & 41.90 + 4490.42 & 0.0550 & 0.0572 & & o ii & 4491.23 & v86a & 3d 2p & 4f d3 * & 4 & 6 & 32.00 + 4493.68 & 0.0030 & 0.0031 & & n iii & 4494.52 & & 4p & 4p * & 4 & 4 & 42.31 + 4495.52 & 0.0040 & 0.0042 & & ne ii & 4496.24 & v58e & 3d 2f & 4f 2@xmath405@xmath41 * & 8 & 10 & 27.50 + 4498.12 & 0.0130 & 0.0135 & & ne ii & 4498.92 & v64c & 3d 4p & 4f 2@xmath401@xmath41 * & 2 & 4 & 8.46 + 4498.25 & 0.0030 & 0.0031 & & ne ii & 4499.12 & v64c & 3d 4p & 4f 2@xmath401@xmath41 * & 2 & 2 & 43.33 + 4502.14 & 0.0030 & 0.0031 & & ar ii & 4503.09 & v63 & 4p 2d * & 5s 4p & 4 & 6 & 10.00 + 4507.60 & 0.0180 & 0.0187 & & ne ii & 4508.13 & v69a & 3d 2p & 4f 1@xmath402@xmath41 * & 2 & 4 & 16.11 + 4510.07 & 0.1610 & 0.1672 & & n iii & 4510.91 & v3 & 3s 4p * & 3p 4d & 4 & 6 & 3.73 + & & & * & n iii & 4510.91 & v3 & 3s 4p * & 3p 4d & 2 & 4 & + & & & * & [ k iv ] & 4510.92 & & 3p4 1d & 3p4 1s & 5 & 1 & + 4514.02 & 0.0460 & 0.0477 & & n iii & 4514.86 & v3 & 3s 4p * & 3p 4d & 6 & 8 & 6.52 + & & & * & ne ii & 4514.88 & v58b & 3d 2f & 4f 2@xmath403@xmath41 * & 8 & 8 & + 4515.93 & 0.0110 & 0.0114 & & ne ii & 4516.66 & v64d & 3d 4p & 4f 2@xmath402@xmath41 * & 2 & 4 & 29.09 + & & & * & ne ii & 4517.40 & v58a & 3d 2f & 4f 2@xmath404@xmath41 * & 8 & 8 & + & & & * & ne ii & 4517.83 & v58a & 3d 2f & 4f 2@xmath404@xmath41 * & 8 & 10 & + 4517.31 & 0.0430 & 0.0446 & & n iii & 4518.15 & v3 & 3s 4p * & 3p 4d & 2 & 2 & 7.44 + & & & * & ne ii & 4522.31 & v55d & 3d 4f & 4f 2@xmath402@xmath41 * & 6 & 6 & + & & & * & ne ii & 4522.52 & v55d & 3d 4f & 4f 2@xmath402@xmath41 * & 6 & 4 & + 4522.74 & 0.0437 & 0.0453 & & n iii & 4523.58 & v3 & 3s 4p * & 3p 4d & 4 & 4 & 6.08 + 4529.60 & 0.0450 & 0.0466 & & n ii & 4530.41 & v58b & 3d 1f * & 4f 2@xmath405@xmath41 & 7 & 9 & 14.89 + 4530.05 & 0.0060 & 0.0062 & & n iii & 4530.86 & v3 & 3s 4p * & 3p 4d & 4 & 2 & : + & & & * & o iii & 4534.31 & v48 & 4f d@xmath403@xmath41 & 5 g g@xmath404@xmath41 * & 7 & 16 & + & & & * & ne ii & 4534.52 & v55b & 3d 4f & 4f 2@xmath403@xmath41 * & 6 & 6 & + 4533.77 & 0.0650 & 0.0673 & & n iii & 4534.58 & v3 & 3s 4p * & 3p 4d & 6 & 6 & 9.38 + & & & * & ne ii & 4534.64 & v55b & 3d 4f & 4f 2@xmath403@xmath41 * & 6 & 8 & + & & & * & ne ii & 4535.37 & v55c & 3d 4f & 4f 2@xmath401@xmath41 * & 4 & 4 & + & & & * & ne ii & 4535.57 & v55c & 3d 4f & 4f 2@xmath401@xmath41 * & 4 & 2 & + & & & * & ne ii & 4537.20 & v55a & 3d 4f & 4f 2@xmath404@xmath41 * & 6 & 8 & + & & & * & o iii & 4538.99 & v48 & 4f d@xmath403@xmath41 & 5 g g@xmath404@xmath41 * & 5 & 7 & + 4538.90 & 0.0240 & 0.0248 & & n iii & 4539.71 & v12 & 4p 2p * & 5s 2s & 2 & 2 & 14.17 + 4540.78 & 0.5490 & 0.5680 & & he ii & 4541.59 & 4.9 & 4f+ 2f * & 9g+ 2 g & 32 & & 1.02 + & & & * & ne ii & 4544.12 & v64c & 3d 4p & 4f 2@xmath401@xmath41 * & 4 & 4 & + 4544.04 & 0.0480 & 0.0496 & & n iii & 4544.85 & v12 & 4p 2p * & 5s 2s & 4 & 2 & 6.25 + 4546.49 & 0.0052 & 0.0054 & & n iii & 4547.30 & v3 & 3s 4p * & 3p 4d & 6 & 4 & 56.67 + 4551.72 & 0.0320 & 0.0330 & & n
ii & 4552.53 & v58a & 3d 1f * & 4f 2@xmath404@xmath41 & 7 & 9 & 0.97 + 4551.86 & 0.0170 & 0.0176 & & si iii & 4552.62 & v2 & 4s 3s & 4p 3p * & 3 & 5 & 1.18 + & & & * & ne ii & 4553.17 & v55d & 3d 4f & 4f 2@xmath402@xmath41 * & 4 & 6 & + & & & * & ne ii & 4553.40 & v55d & 3d 4f & 4f 2@xmath402@xmath41 * & 4 & 4 & + 4553.29 & 0.0007 & 0.0007 & & n ii & 4554.10 & v58a & 3d 1f * & 4f 2@xmath404@xmath41 & 7 & 7 & : + 4557.61 & 0.0030 & 0.0031 & & fe ii ] & 4558.57 & v20 & a2d & z6p * & 4 & 6 & 33.33 + 4561.27 & 0.0070 & 0.0072 & & ne ii & 4561.99 & v64d & 3d 4p & 4f 2@xmath402@xmath41 * & 4 & 6 & 71.43 + 4561.81 & 0.0250 & 0.0258 & & mg i ] & 4562.60 & & 3s2 1s & 3s3p 3p * & 1 & 5 & 32.00 + 4564.18 & 0.0030 & 0.0031 & & [ mn v ] & 4564.70 & & 3d3 2f & 3d3 2d1 & 6 & 6 & 33.33 + 4567.06 & 0.0110 & 0.0113 & & si iii & 4567.82 & v2 & 4s 3s & 4p 3p * & 3 & 3 & 13.64 + 4568.29 & 0.0140 & 0.0144 & & ne ii & 4569.05 & v69b & 3d 2p & 4f 1@xmath403@xmath41 * & 4 & 6 & 22.86 + 4568.85 & 0.0020 & 0.0021 & & ne ii & 4569.60 & v58d & 3d 2f & 4f 2@xmath402@xmath41 * & 6 & 6 & : + 4570.23 & 0.0840 & 0.0866 & & mg i ] & 4571.10 & & 3s2 1s & 3s3p 3p * & 1 & 3 & 2.98 + 4573.61 & 0.0030 & 0.0031 & & ne ii & 4574.42 & v64b & 3d 4p & 4f 2@xmath403@xmath41 * & 4 & 6 & : + 4574.15 & 0.0040 & 0.0041 & & si iii & 4574.76 & v2 & 4s 3s & 4p 3p * & 3 & 1 & 12.50 + 4574.48 & 0.0020 & 0.0021 & & mg i ] & 4575.29 & & 3s2 1s & 3s3p 3p * & 1 & 1 & : + 4575.58 & 0.0020 & 0.0021 & & [ fe ii ] & 4576.39 & & 3d7 2 g & ( 1d2)4s 2d & 8 & 4 & 45.00 + 4578.66 & 0.0020 & 0.0021 & & n ii & 4579.71 & & ? & ? & & & 26.32 + 4579.81 & 0.0030 & 0.0031 & & [ cr ii ] & 4580.80 & & a6s & a4p & 6 & 4 & 24.14 + 4580.82 & 0.0008 & 0.0008 & & si ii & 4581.86 & & 2d & 2p * & 6 & 4 & : + 4584.01 & 0.0020 & 0.0021 & & ne ii & 4584.77 & v58a & 3d 2f & 4f 2@xmath404@xmath41 * & 6 & 8 & : + 4586.85 & 0.0040 & 0.0041 & & o iii & 4587.72 & & 4p 3p & 5s 3p * & 3 & 5 & 80.00 + 4587.39 & 0.0060 & 0.0062 & & ne ii & 4588.14 & v69a & 3d 2p & 4f 1@xmath402@xmath41 * & 4 & 6 & 55.00 + 4590.22 & 0.0850 & 0.0874 & & o ii & 4590.97 & v15 & 3s 2d & 3p 2f * & 6 & 8 & 6.46 + 4592.55 & 0.0060 & 0.0062 & & c iii & 4593.30 & v51 & 4f 3d & 5g 3f * & 7 & 9 & 28.33 + & & & * & o ii & 4595.96 & v15 & 3s 2d & 3p 2f * & 6 & 6 & 15.00 + 4595.43 & 0.0603 & 0.0621 & & o ii & 4596.18 & v15 & 3s 2d & 3p 2f * & 4 & 6 & 2.81 + 4599.41 & 0.0070 & 0.0072 & & ne ii & 4600.16 & v64d & 3d 4p & 4f 2@xmath402@xmath41 * & 6 & 6 & 28.57 + 4600.73 & 0.0160 & 0.0164 & & n ii & 4601.48 & v5 & 3s 3p * & 3p 3p & 3 & 5 & 21.88 + 4601.38 & 0.0503 & 0.0517 & & o ii & 4602.13 & v92b & 3d 2d & 4f f3 * & 4 & 6 & 7.02 + 4606.51 & 0.0350 & 0.0359 & & [ fe iii ] & 4607.03 & & 3d6 5d & 3d6 3f & 9 & 7 & 4.86 + & & & * & n ii & 4607.16 & v5 & 3s 3p * & 3p 3p & 1 & 3 & + 4608.54 & 0.1250 & 0.1283 & & o ii & 4609.44 & v92a & 3d 2d & 4f f4 * & 6 & 8 & 3.28 + 4609.30 & 0.0400 & 0.0411 & & o ii & 4610.20 & v92c & 3d 2d & 4f f2 * & 4 & 6 & 13.25 + 4610.17 & 0.0120 & 0.0123 & & o ii & 4611.07 & v92c & 3d 2d & 4f f2 * & 4 & 4 & 16.67 + 4612.02 & 0.0040 & 0.0041 & & ne ii & 4612.93 & v64b & 3d 4p & 4f 2@xmath403@xmath41 * & 6 & 8 & : + 4612.24 & 0.0050 & 0.0051 & & o ii & 4613.14 & v92b & 3d 2d & 4f f3 * & 6 & 6 & : + 4612.78 & 0.0110 & 0.0113 & & o ii & 4613.68 & v92b & 3d 2d & 4f f3 * & 6 & 8 & 87.27 + 4612.96 & 0.0020 & 0.0021 & & n ii & 4613.87 & v5 & 3s 3p * & 3p 3p & 3 & 3 & : + 4614.66 & 0.0050 & 0.0051 & & ne ii & 4615.56 & v64a & 3d 4p & 4f 2@xmath404@xmath41 * & 6 & 8 & 20.00 + 4615.19 & 0.0070 & 0.0072 & & ne ii & 4616.09 & v67b & 3d 2p & 4f 2@xmath401@xmath41 * & 2 & 4 & 15.71 + 4617.38 & 0.0020 & 0.0021 & & c ii & 4618.40 & v50 & 3d 2f * & 4f 2 g & 6 & 8 & : + & & & * & ni ii & 4618.39 & & 4fsp4f * & 3p 5d4f & 6 & 6 & + 4618.21 & 0.0020 & 0.0021 & & c ii & 4619.23 & v50 & 3d 2f * & 4f 2 g & 8 & 10 & : + 4619.13 & 0.0120 & 0.0123 & & c ii & 4620.11 & & ? & ? & & & 42.37 + 4620.25 & 0.0190 & 0.0195 & & o ii & 4621.27 & v92c & 3d 2d & 4f f2 * & 6 & 6 & 69.47 + & & & * & n ii & 4621.39 & v5 & 3s 3p * & 3p 3p & 3 & 1 & + 4621.12 & 0.0003 & 0.0003 & & o ii & 4622.14 & v92c & 3d 2d & 4f f2 * & 6 & 4 & : + 4627.14 & 0.0090 & 0.0092 & & c ii & 4627.44 & v50 & 3d 2f * & 4f 2 g & 8 & 8 & : + & & & * & ne ii & 4627.98 & v73c & 4s 4p & 4f 1@xmath403@xmath41 * & 6 & 8 & + 4629.70 & 0.0650 & 0.0666 & & n ii & 4630.54 & v5 & 3s 3p * & 3p 3p & 5 & 5 & 20.77 + 4633.30 & 1.1070 & 1.1334 & & n iii & 4634.14 & v2 & 3p 2p * & 3d 2d & 2 & 4 & 1.61 + & & & * & c ii & 4637.63 & v12.01 & 4p 2p * & 6d 2d & 2 & 4 & + 4638.02 & 0.3270 & 0.3346 & & o ii & 4638.86 & v1 & 3s 4p & 3p 4d * & 2 & 4 & 4.92 + & & & * & c ii & 4638.91 & v12.01 & 4p 2p * & 6d 2d & 4 & 6 & + & & & * & c ii & 4639.07 & v12.01 & 4p 2p * & 6d 2d & 4 & 4 & + 4639.80 & 2.2110 & 2.2621 & & n iii & 4640.64 & v2 & 3p 2p * & 3d 2d & 4 & 6 & 1.29 + 4640.97 & 0.4270 & 0.4368 & & o ii & 4641.81 & v1 & 3s 4p & 3p 4d * & 4 & 6 & 16.93 + 4641.01 & 0.2210 & 0.2261 & & n iii & 4641.85 & v2 & 3p 2p * & 3d 2d & 4 & 4 & 1.63 + & & & * & n ii & 4643.09 & v5 & 3s 3p * & 3p 3p & 5 & 3 & + 4646.74 & 0.1640 & 0.1677 & & c iii & 4647.42 & v1 & 3s 3s & 3p 3p * & 3 & 5 & 2.62 + 4648.34 & 0.6520 & 0.6664 & & o ii & 4649.13 & v1 & 3s 4p & 3p 4d * & 6 & 8 & 1.33 + 4649.57 & 0.0990 & 0.1012 & & c iii & 4650.25 & v1 & 3s 3s & 3p 3p * & 3 & 3 & 2.63 + 4650.05 & 0.1715 & 0.1753 & & o ii & 4650.84 & v1 & 3s 4p & 3p 4d * & 2 & 2 & 1.27 + 4650.79 & 0.0330 & 0.0337 & & c iii & 4651.47 & v1 & 3s 3s & 3p 3p * & 3 & 1 & 2.73 + 4657.55 & 0.1430 & 0.1460 & & c iv & 4658.30 & v8 & 5 g 2 g & 6h+ 2h * & & & 40.84 + & & & * & [ fe iii ] & 4658.05 & f3 & 3d6 5d & 3d6 3f2 & 9 & 9 & + 4660.82 & 0.2130 & 0.2174 & & o ii & 4661.63 & v1 & 3s 4p & 3p 4d * & 4 & 4 & 6.43 + 4665.06 & 0.0040 & 0.0041 & & c iii & 4665.86 & v5 & 3s 3p * & 3p 3p & 5 & 5 & : + 4666.21 & 0.0070 & 0.0071 & & [ fe iii ] & 4667.01 & & 3d6 5d & 3d6 3f2 & 7 & 5 & 50.00 + 4668.32 & 0.0058 & 0.0059 & & o ii & 4669.27 & v89b & 3d 2d & 4f d2 * & 4 & 6 & : + 4668.32 & 0.0085 & 0.0087 & & [ p ii ] & 4669.25 & & 3p2 3p & 3p2 1s & 3 & 1 & 40.5 + 4672.98 & 0.0510 & 0.0520 & & o ii & 4673.73 & v1 & 3s 4p & 3p 4d * & 4 & 2 & 7.65 + 4675.49 & 0.1550 & 0.1580 & & o ii & 4676.24 & v1 & 3s 4p & 3p 4d * & 6 & 6 & 3.94 + 4676.32 & 0.0090 & 0.0092 & & o ii & 4677.07 & v91 & 3d 2d & 4f g4 * & 6 & 8 & 1.11 + 4677.39 & 0.0120 & 0.0122 & & n ii & 4678.14 & v61b & 3d 1p * & 4f 2@xmath402@xmath41 & 3 & 5 & : + 4679.83 & 0.0020 & 0.0020 & & o ii & 4680.58 & v89b & 3d 2d & 4f d2 * & 6 & 6 & : + 4680.02 & 0.0006 & 0.0006 & & o ii & 4680.77 & v89b & 3d 2d & 4f d2 * & 6 & 4 & : + 4684.93 & 15.1650 & 15.4396 & & he ii & 4685.68 & 3.4 & 3d+ 2d & 4f+ 2f * & 18 & 32 & 1.22 + 4693.89 & 0.0179 & 0.0182 & & n ii & 4694.64 & v61a & 3d 1p * & 4f 2@xmath403@xmath41 & 3 & 5 & 11.50 + 4695.60 & 0.0150 & 0.0153 & & o ii & 4696.35 & v1 & 3s 4p & 3p 4d * & 6 & 4 & 46.00 + 4695.60 & 0.0090 & 0.0092 & & o ii & 4696.35 & v89a & 3d 2d & 4f d3 * & 4 & 6 & : + 4698.47 & 0.0240 & 0.0244 & & o ii & 4699.22 & v25 & 3p 2d * & 3d 2f & 4 & 6 & 38.75 + 4700.78 & 0.0450 & 0.0457 & & [ fe iii ] & 4701.53 & f3 & 3d6 5d & 3d6 3f2 & 7 & 7 & 20.00 + 4704.60 & 0.0210 & 0.0213 & & o ii & 4705.35 & v25 & 3p 2d * & 3d 2f & 6 & 8 & 19.52 + 4706.68 & 0.0150 & 0.0152 & & o ii & 4707.48 & v89a & 3d 2d & 4f d3 * & 6 & 8 & 44.00 + 4707.01 & 0.0260 & 0.0264 & & o ii & 4707.81 & v89a & 3d 2d & 4f d3 * & 6 & 6 & 42.31 + 4710.57 & 3.5050 & 3.5587 & & [ ar iv ] & 4711.37 & f1 & 3p3 4s * & 3p3 2d * & 4 & 6 & 1.64 + 4712.35 & 0.5790 & 0.5878 & & he i & 4713.20 & v12 & 2p 3p * & 4s 3s & 9 & 3 & 6.46 + 4713.36 & 0.0520 & 0.0528 & & [ ne iv ] & 4714.17 & & 2p3 2d * & 2p3 2p * & 6 & 2 & : + 4714.86 & 0.0140 & 0.0142 & & [ ne iv ] & 4715.66 & & 2p3 2d * & 2p3 2p * & 6 & 4 & : + 4717.96 & 0.0030 & 0.0030 & & o ii & 4718.89 & & 4d 4d & 4fd@xmath402@xmath41 * & 4 & 6 & : + 4723.36 & 0.0230 & 0.0233 & & [ ne iv ] & 4724.17 & f1 & 2p3 2d * & 2p3 2p * & 4 & 4 & 43.04 + 4724.87 & 0.0190 & 0.0193 & & [ ne iv ] & 4725.67 & f1 & 2p3 2d * & 2p3 2p * & 4 & 2 & 52.11 + 4733.18 & 0.0210 & 0.0213 & & [ fe iii ] & 4733.91 & & 3d6 5d & 3d6 3f2 & 5 & 5 & 21.90 + 4739.44 & 3.9700 & 4.0186 & & [ ar iv ] & 4740.17 & f1 & 3p3 4s * & 3p3 2d * & 4 & 4 & 1.31 + 4740.98 & 0.0070 & 0.0071 & & o ii & 4741.71 & v25 & 3p 2d * & 3d 2f & 6 & 6 & : + 4750.58 & 0.0070 & 0.0071 & & o ii & 4751.27 & v24 & 3p 2d * & 3d 4d & 6 & 8 & : + 4751.99 & 0.0060 & 0.0061 & & o ii & 4752.68 & v24 & & & & & : + & & & * & ni ii & 4753.06 & & 4f sp4f * & 3p 4d4f & 8 & 10 & + 4754.00 & 0.0240 & 0.0243 & & [ fe iii ] & 4754.69 & & 3d6 5d & 3d6 3f2 & 7 & 9 & 19.58 + 4755.73 & 0.0090 & 0.0091 & & o i & 4756.70 & & ? & ? & & & 11.40 + 4759.93 & 0.0060 & 0.0061 & & o i & 4760.87 & & ? & ? & & & 10.71 + 4761.53 & 0.0030 & 0.0030 & & c i & 4762.31 & v6 & 3s 3p * & 4p 3p & 1 & 3 & 3.33 + 4761.76 & 0.0150 & 0.0151 &
& c i & 4762.54 & v6 & 3s 3p * & 4p 3p & 3 & 5 & 2.67 + 4764.08 & 0.0170 & 0.0172 & & ar ii & 4764.86 & v15 & 4s 2p & 4p 2p * & 2 & 4 & 4.12 + 4765.90 & 0.0050 & 0.0050 & & c i & 4766.68 & v6 & 3s 3p * & 4p 3p & 3 & 3 & 2.00 + 4766.01 & 0.0050 & 0.0050 & & fe i & 4766.79 & & d3f & 3hsp 3 g * & 9 & 11 & 12.00 + 4767.75 & 0.0090 & 0.0091 & & ni ii & 4768.81 & & 3f 4d4d & 4f 2@xmath401@xmath41 * & 4 & 4 & 23.33 + 4768.78 & 0.0180 & 0.0182 & & [ fe iii ] & 4769.43 & & 3d6 5d & 3d6 3f2 & 5 & 7 & 31.67 + 4769.25 & 0.0070 & 0.0071 & & c i & 4770.03 & v6 & 3s 3p * & 4p 3p & 3 & 1 & 71.43 + 4770.49 & 0.0050 & 0.0050 & & o i ] & 4771.63 & & 3p 5p & 7d 3d * & 7 & 3 & 21.82 + 4770.97 & 0.0040 & 0.0040 & & c i & 4771.75 & v6 & 3s 3p * & 4p 3p & 5 & 5 & 2.50 + 4772.15 & 0.0110 & 0.0111 & & ne ii & 4772.93 & & 4p 4d * & 5d 4f & 6 & 8 & 6.36 + 4773.46 & 0.0080 & 0.0081 & & n ii & 4774.24 & v20 & 3p 3d & 3d 3d * & 3 & 5 & 7.50 + 4775.13 & 0.0030 & 0.0030 & & c i & 4775.91 & v6 & 3s 3p * & 4p 3p & 5 & 3 & 30.00 + 4776.90 & 0.0090 & 0.0091 & & [ fe iii ] & 4777.68 & & 3d6 5d & 3d6 3f4 & 3 & 5 & 13.33 + & & & * & o
ii & 4777.79 & & 4d 4p & 4f d2 * & 4 & 6 & + 4778.94 & 0.0040 & 0.0040 & & n ii & 4779.72 & v20 & 3p 3d & 3d 3d * & 3 & 3 & 22.50 + 4780.41 & 0.0040 & 0.0040 & & n ii & 4781.19 & v20 & 3p 3d & 3d 3d * & 5 & 7 & 27.50 + 4781.75 & 0.0050 & 0.0050 & & ne ii & 4782.12 & v71d & 4s 4p & 4f 2@xmath402@xmath41 * & 6 & 6 & 20.00 + 4783.46 & 0.0030 & 0.0030 & & fe ii & 4784.43 & & e6f & 4f 1@xmath404@xmath41 * & 10 & 10 & 30.00 + 4784.81 & 0.0040 & 0.0040 & & ni ii & 4785.86 & & 4fsp4 g * & 3f 7d4f & 10 & 8 & 12.50 + 4787.60 & 0.0240 & 0.0242 & & [ cr v ] & 4788.51 & & 3d2 3f & 3d2 1 g & 9 & 9 & 4.17 + & & & * & n ii & 4788.13 & v20 & 3p 3d & 3d 3d * & 5 & 5 & + 4788.70 & 0.0160 & 0.0161 & & [ f ii ] & 4789.45 & & 2p4 3p & 2p4 1d & 5 & 5 & : + 4789.65 & 0.0170 & 0.0171 & & o i & 4790.72 & & 3p & 3d * & 5 & 7 & 17.65 + 4791.10 & 0.0180 & 0.0181 & & o i & 4792.12 & & 3p & 3d * & 5 & 3 & 67.22 + 4792.19 & 0.0120 & 0.0121 & & [ co ii ] & 4793.12 & & a5f & c3p & 9 & 5 & 80.00 + 4792.79 & 0.0030 & 0.0030 & & n ii & 4793.65 & v20 & 3p 3d & 3d 3d * & 5 & 3 & 16.67 + & & & * & ni ii & 4793.83 & & 4fsp4 g * & 3f 7d4f & 10 & 10 & + 4793.59 & 0.0230 & 0.0232 & & fe ii ] & 4794.49 & & e6f & 4f 1@xmath403@xmath41 * & 8 & 6 & 57.39 + 4795.27 & 0.0260 & 0.0262 & & c i & 4796.08 & v18.04 & 2p3 3d * & 5p 3d & 5 & 3 & 7.69 + 4797.47 & 0.0190 & 0.0191 & & [ fe ii ] & 4798.27 & & 4s 6d & 4s 4p & 2 & 4 & 11.58 + 4798.06 & 0.0260 & 0.0262 & & [ fe iii ] & 4799.45 & & 5d & 3f4 & 1 & 5 & 9.23 + 4799.62 & 0.0260 & 0.0262 & & n i & 4800.70 & & 3p 4d * & 8d 4p & 4 & 2 & 6.15 + 4801.53 & 0.0330 & 0.0332 & & c ii & 4802.70 & v17.08 & 4f 2f * & 8 g 2 g & 22 & 26 & 8.18 + 4802.43 & 0.0320 & 0.0322 & & n ii & 4803.29 & v20 & 3p 3d & 3d 3d * & 7 & 7 & 15.31 + 4804.04 & 0.0100 & 0.0101 & & [ fe ii ] & 4805.02 & & a6d & a2h & 10 & 10 & 10.00 + 4806.12 & 0.0090 & 0.0090 & & ni i & 4806.99 & v163 & z3d * & f3f & 7 & 9 & 7.78 + 4809.42 & 0.0012 & 0.0012 & & n ii & 4810.31 & v20 & 3p 3d & 3d 3d * & 7 & 5 & : + 4813.70 & 0.0030 & 0.0030 & & [ fe ii ] & 4814.53 & & 3d7 4f & 4s 4f & 10 & 10 & 26.67 + 4814.83 & 0.0150 & 0.0151 & & n ii & 4815.62 & & 3p 5d * & 3d 5p & 3 & 3 & 8.00 + 4818.92 & 0.0020 & 0.0020 & & si iii & 4819.72 & v9 & 4f 3f * & 5
g 3 g & 7 & 9 & 45.00 + 4820.70 & 0.0050 & 0.0050 & & fe iii & 4821.61 & & 4p 5s5p & 4d 5p5d * & 5 & 5 & 20.00 + 4822.33 & 0.0080 & 0.0080 & & c i & 4823.24 & & 3p 3d & 4d 3f * & 5 & 7 & 12.50 + 4824.06 & 0.0050 & 0.0050 & & ni ii & 4825.05 & & 3p 4d4f & 6f 2@xmath404@xmath41 * & 8 & 10 & 30.00 + 4825.48 & 0.0120 & 0.0120 & & fe ii & 4826.38 & & 4s2 2 g & 1 g 4p2f * & 8 & 8 & 14.17 + 4828.17 & 0.0050 & 0.0050 & & si iii & 4828.97 & v9 & 4f 3f * & 5 g 3 g & 9 & 11 & 22.00 + 4828.31 & 0.0002 & 0.0002 & & si iii & 4829.11 & v9 & 4f 3f * & 5 g 3 g & 9 & 9 & : + 4830.14 & 0.0120 & 0.0120 & & c i & 4831.02 & & ? & ? & & & 8.33 + 4832.74 & 0.0040 & 0.0040 & & fe ii & 4833.19 & v30 & a4h & z6f * & 12 & 10 & 45.00 + 4834.09 & 0.0040 & 0.0040 & & ni ii & 4835.11 & & 3p 4d4d & 4f``2@xmath402@xmath41 * & 8 & 6 & 30.00 + 4842.76 & 0.0210 & 0.0210 & & o ii & 4843.37 & v105 & 3d 2s & 4f p1 * & 4 & 6 & : + 4846.41 & 0.0160 & 0.0160 & & ne ii & 4847.23 & & 4d 4p & 8f 2@xmath403@xmath41 * & 6 & 8 & : + 4850.29 & 0.0130 & 0.0130 & & ni ii & 4851.32 & & 3f 4d4d & 3p 5p4p * & 6 & 4 & : + 4851.94 & 0.0290 & 0.0290 & & [ fe ii ] & 4852.73 & & 3d7 4f & 4s 4f & 8 & 4 & : + 4858.03 & 2.9670 & 2.9673 & & n iii & 4858.82 & v9 & 3p 4d & 3d 4f * & 6 & 10 & 12.60 + 4858.53 & 0.0870 & 0.0870 & & he ii & 4859.32 & 4.8 & 4f+ 2f * & 8g+ 2 g & 32 & & 19.66 + 4860.48 & 0.0820 & 0.0820 & & n iii & 4861.27 & v9 & 3p 4d & 3d 4f * & 6 & 8 & 19.63 + 4860.61 & 100.0000 & 100.0000 & & h i & 4861.32 & h4 & 2p+ 2p * & 4d+ 2d & 8 & 32 & 0.00 + 4866.17 & 0.0570 & 0.0570 & & n iii & 4867.15 & v9 & 3p 4d & 3d 4f * & 12 & 14 & 5.96 + 4872.65 & 0.0140 & 0.0140 & & n iii & 4873.57 & v9 & 3p 4d & 3d 4f * & 6 & 6 & 6.43 + 4873.70 & 0.0100 & 0.0100 & & [ fe ii ] & 4874.49 & & 3d7 4f & 4s 4f & 8 & 6 & : + 4880.30 & 0.0630 & 0.0629 & & [ fe iii ] & 4881.00 & f2 & 3d6 5d & 3d6 3h & 9 & 9 & 15.87 + 4880.91 & 0.0570 & 0.0569 & & n iii & 4881.81 & v9 & 3p 4d & 3d 4f * & 6 & 4 & 1.75 + 4883.40 & 0.0090 & 0.0090 & & n iii & 4884.13 & v9 & 3p 4d & 3d 4f * & 8 & 8 & 12.22 + 4885.31 & 0.0030 & 0.0030 & & [ co v ] & 4886.23 & & 4p & 4f & 6 & 4 & 43.33 + 4890.13 & 0.0130 & 0.0130 & & o ii & 4890.86 & v28 & 3p 4s * & 3d 4p & 4 & 2 & 4.62 + 4893.06 & 0.0013 & 0.0013 & & [ cr iv ] & 4893.85 & & 3d3 4f & 3d3 2d2 & 6 & 6 & 70.00 + 4894.41 & 0.0020 & 0.0020 & & n ii & 4895.11 & v1 & 2p3 1d * & 3p 1p & 5 & 3 & 30.00 + 4895.90 & 0.0040 & 0.0040 & & n iii & 4896.63 & v9 & 3p 4d & 3d 4f * & 8 & 6 & 32.50 + 4896.54 & 0.0020 & 0.0020 & & [ cr iv ] & 4897.33 & & 3d3 4f & 3d3 2d2 & 6 & 4 & 65.00 + 4897.82 & 0.0050 & 0.0050 & & [ fe ii ] & 4898.61 & & 3d7 2 g & ( 3d)4s 2d & 10 & 6 & 24.00 + 4898.90 & 0.0100 & 0.0100 & & [ fe iv ] & 4899.97 & & 3d5 4 g & 3d5 4f & 10 & 8 & 8.00 + 4900.91 & 0.0070 & 0.0070 & & o i & 4902.03 & & 3p & 3d * & 5 & 3 & 24.29 + 4902.08 & 0.0280 & 0.0279 & & [ fe iv ] & 4903.07 & & 3d5 4 g & 3d5 4f & 8 & 8 & 5.00 + 4903.54 & 0.0060 & 0.0060 & & o ii & 4904.45 & & 4f d3 * & 3d '' 2d & 8 & 6 & 15.00 + 4904.55 & 0.0060 & 0.0060 & & [ fe ii ] & 4905.34 & & 3d7 4f & 4s 4f & 8 & 8 & 1.67 + 4906.03 & 0.0460 & 0.0458 & & o ii & 4906.83 & v28 & 3p 4s * & 3d 4p & 4 & 4 & 1.74 + 4906.42 & 0.0040 & 0.0040 & & n iii & 4907.20 & v25 & 5d 2d & 7f 2f * & & & 2.50 + 4910.08 & 0.0013 & 0.0013 & & cr ii & 4911.27 & v199 & c4f & w4d * & 4 & 2 & 90.00 + 4911.90 & 0.0030 & 0.0030 & & mn ii & 4912.99 & & d3d & w3f * & 7 & 7 & 36.67 + 4917.22 & 0.0050 & 0.0050 & & [ fe iv ] & 4917.97 & & 3d5 4 g & 3d5 4f & 10 & 10 & 2.00 + 4920.47 & 0.1300 & 0.1292 & & [ fe iii ] & 4921.66 & & 3p2 & 3 g & 5 & 7 & : + 4921.23 & 1.1760 & 1.1689 & & he i & 4921.93 & v48 & 2p 1p * & 4d 1d & 3 & 5 & 9.95 + 4923.67 & 0.0750 & 0.0745 & & o ii & 4924.53 & v28 & 3p 4s * & 3d 4p & 4 & 5 & 7.47 + 4929.85 & 0.0360 & 0.0358 & & [ fe iii ] & 4930.54 & & 3d6 5d & 3d6 3p2 & 3 & 1 & 85.28 + 4930.60 & 0.1230 & 0.1221 & & [ o iii ] & 4931.23 & f1 & 2p2 3p & 2p2 1d & 1 & 5 & 19.27 + 4932.37 & 0.0140 & 0.0139 & & ne ii & 4933.18 & & 2f & 2@xmath403@xmath41 * & 8 & 6 & 70.71 + 4933.88 & 0.0180 & 0.0179 & & n iii & 4934.89 & & ? & ? & & & : + 4935.07 & 0.0210 & 0.0208 & & [ fe iii ] & 4936.36 & & 5d & 3p4 & 7 & 3 & 83.33 + 4936.43 & 0.0180 & 0.0179 & & n i & 4937.43 & v41 & 3p 4d * & 7d 4p & 2 & 4 & : + 4937.50 & 0.0060 & 0.0060 & & fe ii & 4938.50 & & 5d 4d6d & 4f 1@xmath403@xmath41 * & 10 & 8 & : + 4938.83 & 0.0060 & 0.0060 & & ne ii & 4939.66 & & ?
& ? & & & : + 4940.48 & 0.0100 & 0.0099 & & o ii & 4941.07 & v33 & 3p 2p * & 3d 2d & 2 & 4 & 65.00 + 4942.24 & 0.0120 & 0.0119 & & o ii & 4943.00 & v33 & 3p 2p * & 3d 2d & 4 & 6 & 42.50 + 4957.91 & 391.6830 & 387.9524 & & [ o iii ] & 4958.91 & f1 & 2p2 3p & 2p2 1d & 3 & 5 & 0.03 + 4966.12 & 0.2860 & 0.2831 & & [ fe vi ] & 4967.14 & & 3d3 4f & 3d3 2 g & 8 & 10 & 9.79 + 4970.69 & 0.0080 & 0.0079 & & [ cr iv ] & 4971.48 & & 3d3 4f & 3d3 2d2 & 8 & 6 & 8.75 + & & & * & [ fe vi ] & 4971.71 & & 3d3 2p & 3d3 2f & 2 & 6 & + 4971.45 & 0.1290 & 0.1276 & & [ fe vi ] & 4972.47 & & 3d3 4f & 3d3 2 g & 6 & 8 & 10.08 + 4979.28 & 0.0500 & 0.0494 & & o i & 4980.09 & & 3p 3p & 11s 3s * & 5 & 3 & 10.00 + 4986.35 & 0.0070 & 0.0069 & & n ii & 4987.37 & v24 & 3p 3s & 3d 3p * & 3 & 1 & 14.29 + 5005.61 & 1175.7600 & 1159.2320 & & [ o iii ] & 5006.84 & f1 & 2p2 3p & 2p2 1d & 5 & 5 & 0.07 + 5014.66 & 1.5130 & 1.4907 & & he i & 5015.68 & v4 & 2s 1s & 3p 1p * & 1 & 3 & 30.58 + 5015.37 & 0.2310 & 0.2276 & & n ii & 5016.39 & v19 & 3p 3d & 3d 3f * & 5 & 5 & : + 5024.64 & 0.0590 & 0.0581 & & n ii & 5025.66 & v19 & 3p 3d & 3d 3f * & 7 & 7 & 38.64 + 5029.61 & 0.0400 & 0.0394 & & fe ii & 5030.63 & & e6f & 4f 4@xmath405@xmath41 * & 10 & 10 & 58.75 + 5032.57 & 0.0430 & 0.0423 & & [ fe iv ] & 5033.59 & & 3d5 2f2 & 3d5 4 g & 8 & 6 & 60.23 + 5040.01 & 0.1020 & 0.1003 & & n ii & 5040.72 & v19 & 3p 3d & 3d 3f * & 7 & 5 & 76.18 + 5040.30 & 0.0820 & 0.0806 & & si ii & 5041.03 & v5 & 4p 2p * & 4d 2d & 2 & 4 & 67.80 + 5046.73 & 0.1600 & 0.1572 & & he i & 5047.74 & v47 & 2p 1p * & 4s 1s & 3 & 1 & 14.81 + 5054.97 & 0.0880 & 0.0864 & & si ii & 5055.98 & v5 & 4p 2p * & 4d 2d & 4 & 6 & 4.77 + 5055.29 & 0.0100 & 0.0098 & & si ii & 5056.31 & v5 & 4p 2p * & 4d 2d & 4 & 4 & 5.00 + 5057.83 & 0.0530 & 0.0520 & & n iii & 5058.85 & v27 & 5f 2f * & 7
g 2 g & & & 48.11 + 5061.80 & 0.0420 & 0.0412 & & fe ii & 5062.93 & & e6f & 4f 4@xmath404@xmath41 * & 8 & 10 & 8.81 + 5066.60 & 0.0580 & 0.0569 & & ni iii & 5067.52 & & 4f 4d5f & 4f 5p5 g * & 11 & 13 & 7.59 + 5070.51 & 0.0550 & 0.0539 & & ni ii & 5071.50 & & 3f 4d4p & 4f 3@xmath402@xmath41 * & 4 & 6 & 10.91 + 5073.53 & 0.0570 & 0.0559 & & o i & 5074.77 & & 3d * & 3d & 7 & 7 & 11.93 + 5076.37 & 0.0430 & 0.0421 & & n ii & 5077.47 & v61.19 & 4p 3p & 6s 3p * & 3 & 5 & 16.05 + 5079.81 & 0.0360 & 0.0353 & & ni ii & 5080.96 & & 3f 4d4f & 4f 3@xmath404@xmath41 * & 8 & 8 & 12.50 + 5084.00 & 0.0330 & 0.0323 & & [ fe iii ] & 5084.77 & & 3d6 5d & 3d6 3p2 & 1 & 3 & 10.91 + 5088.73 & 0.0300 & 0.0294 & & c i & 5089.63 & v22.11 & 3p 1p & 7d 1d * & 3 & 5 & 11.67 + 5092.84 & 0.0200 & 0.0196 & & fe ii & 5093.77 & & 5d 4d6d & 4f 3@xmath405@xmath41 * & 10 & 10 & 17.00 + & & & * & ne ii & 5093.98 & & 4d 2d & 7f 1@xmath402@xmath41 * & 4 & 6 & + 5102.72 & 0.0140 & 0.0137 & & si iii & 5104.00 & & 6p 3p * & 8d 3d & 1 & 3 & 17.86 + 5107.13 & 0.0080 & 0.0078 & & [ fe ii ] & 5107.94 & & 3d7 4f & ( 3p2)4s 4p & 6 & 2 & 76.25 + 5109.36 & 0.0060 & 0.0059 & & [ fe iii ] & 5110.40 & & 3f2 & 5d & 5 & 1 & : + 5114.07 & 0.0090 & 0.0088 & & fe ii & 5115.06 & & 5d 4d6d & 4f 2@xmath401@xmath41 * & 2 & 2 & 30.00 + 5117.35 & 0.0050 & 0.0049 & & fe ii & 5118.50 & & x4 g * & e4 g & 8 & 8 & 64.00 + 5119.09 & 0.0270 & 0.0264 & & c ii & 5120.10 & v12 & 4p 2p * & 3p 2p & 2 & 4 & 1.48 + 5120.97 & 0.0110 & 0.0107 & & c ii & 5121.82 & v12 & 4p 2p * & 3p 2p & 4 & 4 & 19.09 + 5124.19 & 0.0050 & 0.0049 & & c ii & 5125.20 & v12 & 4p 2p * & 3p 2p & 2 & 2 & 16.00 + 5125.92 & 0.0060 & 0.0059 & & c ii & 5126.93 & v12 & 4p 2p * & 3p 2p & 4 & 2 & 6.67 + 5129.66 & 0.0450 & 0.0439 & & ne ii & 5130.46 & & 4d 4d & 7f 2@xmath401@xmath41 * & 4 & 4 & 10.00 + 5132.09 & 0.0090 & 0.0088 & & c ii & 5132.94 & v16 & 3s 4p * & 3p 4p & 2 & 4 & 48.89 + 5141.98 & 0.0160 & 0.0156 & & fe i & 5142.93 & v16 & a5f & z5f * & 7 & 9 & 49.38 + 5144.45 & 0.0080 & 0.0078 & & n i & 5145.48 & v35 & 3p 4d * & 6d 4p & 2 & 4 & : + 5146.58 & 0.0520 & 0.0506 & & n iii & 5147.88 & v29 & 5 g 2 g & 7h 2h * & & & 35.19 + 5152.13 & 0.0100 & 0.0097 & & n i & 5153.76 & v35 & 3p 4d * & 6d 4p & 4 & 6 & 27.00 + 5157.48 & 0.0050 & 0.0049 & & [ fe vii ] & 5158.89 & & 3d2 3f & 3d2 3p & 7 & 3 & : + 5158.40 & 0.0170 & 0.0165 & & ne ii & 5159.37 & & 4d 4f & 7f 0@xmath403@xmath41 * & 4 & 6 & : + 5167.40 & 0.0050 & 0.0049 & & n ii & 5168.06 & v70 & 3p 5p * & 3d 5d & 3 & 5 & 10.00 + 5169.51 & 0.0060 & 0.0058 & & n ii & 5170.17 & v70 & 3p 5p * & 3d 5d & 3 & 3 & 11.67 + 5170.79 & 0.0120 & 0.0117 & & n ii & 5171.45 & v70 & 3p 5p * & 3d 5d & 5 & 7 & 11.67 + 5171.69 & 0.0020 & 0.0019 & & [ fe ii ] & 5172.47 & & 3d7 2p & 4s 2d & 2 & 4 & 10.00 + & & & * & n ii & 5172.35 & v66 & 3p 5d * & 3d 5f & 3 & 5 & + 5172.31 & 0.0009 & 0.0009 & & n ii & 5173.39 & v66 & 3p 5d * & 3d 5f & 5 & 7 & 10.00 + 5172.73 & 0.0030 & 0.0029 & & ni iii & 5173.64 & & 4f 4d5p & 4f 5p5d * & 5 & 3 & 13.33 + 5173.80 & 0.0080 & 0.0078 & & n ii & 5174.46 & v70 & 3p 5p * & 3d 5d & 5 & 5 & 11.25 + 5175.23 & 0.0050 & 0.0049 & & [ fe vi ] & 5176.04 & & 3d3 4f & 3d3
2 g & 10 & 10 & 12.00 + 5175.38 & 0.0130 & 0.0126 & & n ii & 5176.56 & v70 & 3p 5p * & 3d 5d & 5 & 3 & 29.23 + 5176.40 & 0.0009 & 0.0009 & & n ii & 5177.06 & v66 & 3p 5d * & 3d 5f & 3 & 3 & 10.00 + 5178.69 & 0.0230 & 0.0223 & & n ii & 5179.35 & v70 & 3p 5p * & 3d 5d & 7 & 9 & 12.17 + 5178.86 & 0.0070 & 0.0068 & & n ii & 5179.52 & v66 & 3p 5d * & 3d 5f & 9 & 11 & 12.86 + 5179.69 & 0.0013 & 0.0013 & & n ii & 5180.35 & v66 & 3p 5d * & 3d 5f & 5 & 5 & 20.00 + 5182.16 & 0.0060 & 0.0058 & & n ii & 5183.20 & v70 & 3p 5p * & 3d 5d & 7 & 7 & 11.67 + 5182.55 & 0.0130 & 0.0126 & & mg i & 5183.60 & v2 & 3p 3p * & 4s 3s & 5 & 3 & 29.23 + 5185.33 & 0.0150 & 0.0146 & & n ii & 5186.20 & v70 & 3p 5p * & 3d 5d & 7 & 5 & 18.00 + 5190.46 & 0.1040 & 0.1009 & & [ ar iii ] & 5191.82 & f3 & 3p4 1d & 3p4 1s & 5 & 1 & 3.94 + 5196.53 & 0.0580 & 0.0562 & &
[ n i ] & 5197.90 & f1 & 2p3 4s * & 2p3 2d * & 4 & 4 & 6.55 + 5198.89 & 0.0350 & 0.0339 & & [
n i ] & 5200.26 & f1 & 2p3 4s * & 2p3 2d * & 4 & 6 & 10.29 + 5217.42 & 0.0080 & 0.0077 & & ne ii & 5218.45 & & 4d 2f & 7f 2@xmath404@xmath41 * & 8 & 10 & 20.00 + 5228.88 & 0.0130 & 0.0126 & & fe i & 5229.79 & & z3f * & f5d & 9 & 9 & 17.69 + 5232.70 & 0.0210 & 0.0203 & & [ fe iv ] & 5233.76 & & 3d5 4 g & 3d5 2f2 & 10 & 8 & 12.38 + 5236.24 & 0.0120 & 0.0116 & & [ fe iv ] & 5237.30 & & 3d5 4 g & 3d5 2f2 & 8 & 8 & 20.83 + 5241.05 & 0.0120 & 0.0116 & & [ cr ii ] & 5241.96 & & a6d & a4f & 6 & 10 & 15.83 + 5246.36 & 0.0080 & 0.0077 & & [ co iii ] & 5247.41 & & 3d7 4f & 3d7 2p & 4 & 2 & 42.50 + 5249.01 & 0.0070 & 0.0068 & & c ii & 5249.90 & v30 & 3d 4f * & 4p 4d & 4 & 6 & 51.43 + 5258.10 & 0.0220 & 0.0212 & & c ii & 5259.06 & v30 & 3d 4f * & 4p 4d & 8 & 6 & 6.36 + 5260.67 & 0.0180 & 0.0174 & & [ fe ii ] & 5261.62 & & 3d7 4f & 3d7 4h & 8 & 12 & 8.33 + 5263.52 & 0.0220 & 0.0212 & & mg ii & 5264.30 & v17 & 4d 2d & 7f 2f * & 10 & 14 & 7.73 + 5266.42 & 0.0320 & 0.0308 & & ne ii & 5267.35 & & 4d 2f & 7f 1@xmath402@xmath41 * & 6 & 6 & 6.88 + 5267.92 & 0.0320 & 0.0308 & & [ fe ii ] & 5268.87 & & 3d7 4f & ( 3p2)4s 4p & 6 & 4 & 6.88 + 5269.45 & 0.0900 & 0.0867 & & [ fe iii ] & 5270.40 & & 3d6 5d & 3d6 3p2 & 7 & 5 & 2.67 + 5272.40 & 0.0180 & 0.0173 & & [ fe ii ] & 5273.35 & & 3d7 4f & ( 3p2)4s 4p & 10 & 6 & 11.67 + 5274.17 & 0.0120 & 0.0116 & & o i & 5275.12 & v27 & 3p 3p & 7d 3d * & 5 & 7 & 13.33 + 5276.84 & 0.0170 & 0.0164 & & [ fe vi ] & 5277.80 & & 3d3 4f & 3d3 4p & 4 & 4 & 19.41 + 5278.72 & 0.0050 & 0.0048 & & [ cr ii ] & 5279.88 & & 3d5 4f & ( 5d)4p 4f * & 10 & 8 & 2.00 + 5279.76 & 0.0110 & 0.0106 & & fe ii & 5280.92 & & e6 g & 4f 1@xmath402@xmath41 * & 4 & 4 & 74.55 + 5282.11 & 0.0100 & 0.0096 & & [ fe ii ] & 5283.11 & & a4d & a2f & 6 & 8 & 31.00 + 5285.32 & 0.0100 & 0.0096 & & c ii & 5286.47 & v56 & 4p 4d & 6s 4p * & 8 & 6 & 15.00 + 5288.45 & 0.0260 & 0.0250 & & o ii & 5289.48 & & 4p * & 4d & 2 & 4 & 11.92 + 5290.51 & 0.0090 & 0.0087 & & fe ii & 5291.67 & & e6 g & 4f 3@xmath406@xmath41 * & 10 & 12 & 38.89 + 5301.76 & 0.0050 & 0.0048 & & si iii & 5302.60 & v32 & 5f 3f * & 7d 3d & 5 & 7 & 28.00 + 5304.41 & 0.0090 & 0.0086 & & c iii & 5305.10 & v46 & 5f 3f * & 7
g 3 g & & & 15.56 + 5307.82 & 0.0020 & 0.0019 & & [ ca v ] & 5309.11 & & 3p4 3p & 3p4 1d & 5 & 5 & 50.00 + 5312.85 & 0.0040 & 0.0038 & & ne ii & 5313.85 & & 4d 2p & 7f 1@xmath402@xmath41 * & 4 & 6 & 22.50 + 5319.44 & 0.0040 & 0.0038 & & o ii & 5320.44 & & f@xmath403@xmath41 * & 2d & 6 & 4 & 35.00 + 5321.62 & 0.0130 & 0.0125 & & [ cl iv ] & 5323.28 & & 3p2 1d & 3p2 1s & 5 & 1 & 9.23 + 5333.72 & 0.0080 & 0.0077 & & cr ii & 5334.87 & v43 & b4f & z4f * & 4 & 4 & 12.50 + 5337.36 & 0.0080 & 0.0077 & & fe i & 5338.34 & & x5f * & 7/2@xmath403/2@xmath41 & 7 & 5 & 13.75 + 5341.25 & 0.0390 & 0.0374 & & c ii & 5342.40 & v17.06 & 4f 2f * & 7 g 2 g & 22 & 26 & 5.13 + 5344.53 & 0.0220 & 0.0211 & & [ kr iv ] & 5346.02 & & 4p3 4s & 4p3 2d & 4 & 6 & 40.45 + 5355.08 & 0.0050 & 0.0048 & & [ fe iii ] & 5355.87 & & 5d & 3p4 & 5 & 5 & : + 5358.11 & 0.0130 & 0.0124 & & ne ii & 5359.17 & & 4d 4f & 7f 2@xmath404@xmath41 * & 8 & 10 & : + & & & * & [ fe ii ] & 5359.24 & & e6 g & 4f 3@xmath403@xmath41 * & 8 & 6 & + 5362.87 & 0.0160 & 0.0153 & & o ii & 5363.53 & & 4f f4 * & 4d 2f & 8 & 8 & : + 5365.87 & 0.0070 & 0.0067 & & [ mn vi ] & 5366.80 & & 3d2 3f & 3d2 3p & 5 & 5 & : + 5366.93 & 0.0080 & 0.0076 & & c ii & 5368.10 & & 2p3 2p * & 7d 2d & 2 & 4 & : + 5369.63 & 0.0190 & 0.0182 & & [ fe vi ] & 5370.29 & & 3d3 4f & 3d3 2 g & 10 & 8 & : + 5375.59 & 0.0250 & 0.0239 & & [ fe ii ] & 5376.45 & & 3d7 4f & ( 3h)4s 4h & 4 & 8 & 66.00 + 5382.15 & 0.0090 & 0.0086 & & o i & 5383.25 & & 3p & 3d * & 3 & 3 & : + 5390.35 & 0.0080 & 0.0076 & & ne ii & 5391.24 & & 4d 4p & 7f 2@xmath401@xmath41 * & 6 & 4 & : + 5395.37 & 0.0110 & 0.0105 & & fe ii & 5396.59 & & e4 g & 4f 1@xmath403@xmath41 * & 8 & 6 & 59.09 + 5410.22 & 1.2140 & 1.1565 & & he ii & 5411.52 & 4.7 & 4f+ 2f * & 7g+ 2 g & 32 & 98 & 2.54 + & & & * & [ fe iii ] & 5411.98 & & 3d6 5d & 3d6 3p2 & 3 & 5 & + 5422.83 & 0.0020 & 0.0019 & & [ fe iii ] & 5423.90 & & 3d6 3d & 3d6 3p2 & 5 & 1 & 20.00 + 5425.66 & 0.0030 & 0.0029 & & [ fe iii ] & 5426.70 & & 3d6 3d & 3d6 3p2 & 3 & 1 & 16.67 + 5428.68 & 0.0040 & 0.0038 & & fe ii & 5429.99 & & e4 g & 4f 1@xmath404@xmath41 * & 8 & 10 & 15.00 + 5431.17 & 0.0050 & 0.0048 & & si i & 5432.04 & & 1p & ( 1/2,3/2 ) * & 3 & 3 & 12.00 + 5436.23 & 0.0030 & 0.0029 & & n ii & 5437.54 & & 5 g * & 5f & 9 & 7 & 26.67 + 5450.73 & 0.0030 & 0.0028 & & n ii & 5452.08 & v29 & 3p 3p & 3d 3p * & 1 & 3 & 40.00 + & & & * & s ii & 5453.83 & v6 & 4s 4p & 4p 4d * & 6 & 8 & + 5452.87 & 0.0060 & 0.0057 & & n ii & 5454.22 & v29 & 3p 3p & 3d 3p * & 3 & 1 & 18.33 + 5456.27 & 0.0030 & 0.0028 & & n i & 5457.44 & & 3p 2d * & 7d 2d & 4 & 6 & 30.00 + 5459.32 & 0.0070 & 0.0066 & & fe ii & 5460.51 & & e4 g & 4f 1@xmath403@xmath41 * & 6 & 6 & 14.29 + & & & * & n ii & 5462.59 & v29 & 3p 3p & 3d 3p * & 3 & 3 & + 5466.52 & 0.0060 & 0.0057 & & fe ii ] & 5467.72 & & x4f * & e6 g & 10 & 12 & 15.00 + & & & * & ni ii & 5467.80 & & 3f 4d4p & 4f 4@xmath401@xmath41 * & 4 & 2 & + 5469.78 & 0.0040 & 0.0038 & & v ii & 5470.87 & & c3f & y3 g * & 7 & 7 & 27.50 + 5473.45 & 0.0030 & 0.0028 & & v ii & 5474.51 & & c3f & y3 g * & 9 & 7 & 20.00 + & & & * & s ii & 5473.62 & v6 & 4s 4p & 4p 4d * & 2 & 2 & + 5477.51 & 0.0009 & 0.0009 & & c ii & 5478.59 & v34 & 3d 4d * & 4p 4d & 8 & 8 & : + & & & * & n ii & 5478.10 & v29 & 3p 3p & 3d 3p * & 3 & 5 & + 5478.98 & 0.0090 & 0.0085 & & n ii & 5480.06 & v29 & 3p 3p & 3d 3p * & 5 & 3 & 34.44 + 5482.27 & 0.0020 & 0.0019 & & c ii & 5483.35 & v34 & 3d 4d * & 4p 4d & 4 & 6 & : + 5483.76 & 0.0040 & 0.0038 & & [ fe vi ] & 5484.84 & & 3d3 4f & 3d3 4p & 6 & 2 & : + 5484.82 & 0.0004 & 0.0004 & & c ii & 5485.90 & v34 & 3d 4d * & 4p 4d & 6 & 6 & : + 5487.39 & 0.0170 & 0.0161 & & c ii & 5488.47 & v34 & 3d 4d * & 4p 4d & 2 & 4 & 25.88 + 5493.75 & 0.0070 & 0.0066 & & n ii & 5495.67 & v29 & 3p 3p & 3d 3p * & 5 & 5 & 15.71 + 5496.36 & 0.0060 & 0.0057 & & n iii & 5497.20 & & 2d & 2d * & 10 & 10 & 18.33 + 5499.06 & 0.0060 & 0.0057 & & fe ii & 5500.17 & & 3f 4p4 g * & 3 g 4d4h & 12 & 12 & 16.67 + 5506.78 & 0.0040 & 0.0038 & & o iii & 5508.25 & v16 & 3p 1d & 3d 1d * & 5 & 5 & : + & & & * & s ii & 5509.72 & v6 & 4s 4p & 4p 4d * & 4 & 4 & + 5513.10 & 0.0430 & 0.0407 & & mg i & 5514.16 & & 1s & 1p * & 1 & 3 & 20.23 + 5516.42 & 0.4800 & 0.4539 & & [ cl iii ] & 5517.72 & f1 & 3p3 4s * & 3p3 2d * & 4 & 6 & 1.46 + 5524.14 & 0.0090 & 0.0085 & & n ii & 5526.24 & v63 & 3s 5p & 3p 5d * & 3 & 5 & 31.11 + 5528.14 & 0.0120 & 0.0113 & & n ii & 5530.24 & v63 & 3s 5p & 3p 5d * & 5 & 7 & 31.67 + 5533.26 & 0.0480 & 0.0453 & & n ii & 5535.36 & v63 & 3s 5p & 3p 5d * & 7 & 9 & 31.25 + 5536.47 & 0.5790 & 0.5467 & & [ cl iii ] & 5537.89 & f1 & 3p3 4s * & 3p3 2d * & 4 & 4 & 1.35 + 5537.95 & 0.0050 & 0.0047 & & n ii & 5540.06 & v63 & 3s 5p & 3p 5d * & 3 & 1 & 32.00 + 5541.37 & 0.0120 & 0.0113 & & n ii & 5543.47 & v63 & 3s 5p & 3p 5d * & 5 & 5 & 31.67 + 5550.50 & 0.0100 & 0.0094 & & n ii & 5551.92 & v63 & 3s 5p & 3p 5d * & 7 & 7 & 46.00 + 5555.07 & 0.0060 & 0.0057 & & [ fe ii ] & 5556.29 & & 3d7 4f & ( 3p2)4s 4p & 6 & 6 & 60.00 + 5555.07 & 0.0055 & 0.0052 & & s ii & 5556.01 & v6 & 4s 4p & 4p 4d * & 4 & 2 & 65.45 + 5563.31 & 0.0050 & 0.0047 & & n ii & 5565.25 & v63 & 3s 5p & 3p 5d * & 7 & 5 & 72.00 + 5567.50 & 0.0060 & 0.0057 & & fe ii & 5568.60 & & f4d & 4f 4@xmath401@xmath41 * & 2 & 2 & 63.33 + 5576.31 & 0.0380 & 0.0358 & & [ o i ] & 5577.34 & f3 & 2p4 1d & 2p4 1s & 5 & 1 & 11.05 + 5587.04 & 0.0080 & 0.0075 & & fe ii & 5588.23 & & e4 g & 4f 3@xmath405@xmath41 * & 8 & 10 & 21.25 + 5590.96 & 0.0490 & 0.0461 & & o iii & 5592.24 & v5 & 3s 1p * & 3p 1p & 3 & 3 & 5.71 + 5605.13 & 0.0070 & 0.0066 & & [ fe ii ] & 5606.25 & & a4d & b2p & 8 & 4 & 21.43 + 5611.78 & 0.0050 & 0.0047 & & n ii & 5612.34 & & 3d & 3d * & 7 & 5 & 30.00 + 5617.62 & 0.0060 & 0.0056 & & [ ca vii ] & 5618.75 & & 3p2 3p & 3p2 1d & 5 & 5 & 23.33 + 5621.18 & 0.0050 & 0.0047 & & [ mn vi ] & 5622.10 & & 3d2 3f & 3d2 3p & 5 & 1 & 24.00 + 5626.56 & 0.0030 & 0.0028 & & n ii & 5627.76 & & 4p 3p & 5d 3p * & 1 & 3 & 56.67 + 5629.57 & 0.0080 & 0.0075 & & [ fe vi ] & 5631.10 & & 3d3 4f & 3d3 4p & 8 & 4 & 28.75 + 5632.11 & 0.0030 & 0.0028 & & fe ii ] & 5633.37 & & d2f & w4f * & 8 & 10 & 83.33 + 5635.32 & 0.0060 & 0.0056 & & mg i & 5636.53 & & 1s & 1p * & 1 & 3 & 23.33 + 5639.03 & 0.0080 & 0.0075 & & ni ii & 5640.23 & & 1d 4d2p & 5f 2@xmath403@xmath41 * & 4 & 6 & 16.25 + 5643.06 & 0.0050 & 0.0047 & & [ fe ii ] & 5643.97 & & a4f & b4p & 4 & 6 & 20.00 + 5650.37 & 0.0030 & 0.0028 & & fe ii & 5651.63 & & 5d 4d4s & 4f 3@xmath401@xmath41 * & 4 & 4 & 26.67 + 5659.42 & 0.0119 & 0.0110 & & [ cr iii ] & 5660.58 & & 3d & b3p & 5 & 5 & 23.64 + & & & * & [ fe ii ] & 5659.83 & & a4d & a4 g & 6 & 6 & + 5665.20 & 0.0684 & 0.0652 & & n
ii & 5666.63 & v3 & 3s 3p * & 3p 3d & 3 & 5 & 6.64 + 5674.59 & 0.0380 & 0.0355 & & n ii & 5676.02 & v3 & 3s 3p * & 3p 3d & 1 & 3 & 11.95 + 5678.13 & 0.1390 & 0.1355 & & n ii & 5679.56 & v3 & 3s 3p * & 3p 3d & 5 & 7 & 4.78 + 5682.37 & 0.0030 & 0.0028 & & [ fe vi ] & 5683.70 & & 3d3 2d2 & 3d3 2f & 4 & 8 & 73.33 + 5684.89 & 0.0260 & 0.0243 & & n ii & 5686.21 & v3 & 3s 3p * & 3p 3d & 3 & 3 & 11.15 + 5690.67 & 0.0170 & 0.0159 & & [ mn v ] & 5692.00 & & 3d3 4f & 3d3 2 g & 8 & 10 & 15.88 + 5700.47 & 0.0360 & 0.0336 & & [ mn v ] & 5701.82 & & 3d3 4f & 3d3
2 g & 6 & 8 & 7.78 + 5709.44 & 0.0210 & 0.0196 & & n ii & 5710.77 & v3 & 3s 3p * & 3p 3d & 5 & 5 & 12.86 + 5738.15 & 0.0080 & 0.0074 & & si iii & 5739.73 & v4 & 4s 1s & 4p 1p * & 1 & 3 & 13.75 + 5745.67 & 0.0070 & 0.0065 & & [ fe ii ] & 5746.97 & & ( 5d)4s 4d & ( 3p2)4s 2p & 6 & 4 & 4.29 + 5753.07 & 0.4250 & 0.3953 & & [ n ii ] & 5754.64 & f3 & 2p2 1d & 2p2 1s & 5 & 1 & 4.02 + 5755.43 & 0.0003 & 0.0003 & & [ fe ii ] & 5756.74 & & 3d7 2p & ( 3d)4s 2d & 2 & 6 & : + 5765.09 & 0.0070 & 0.0065 & & si i & 5765.97 & & 3d & ( 3/2,5/2 ) * & 7 & 7 & 25.71 + 5770.89 & 0.0050 & 0.0046 & & n i & 5771.96 & v89 & 3p 4s * & 7s 4p & 4 & 4 & 26.00 + 5783.68 & 0.0030 & 0.0028 & & he ii & 5784.94 & 5.40 & 5g+ 2 g & 40h+ 2h * & 50 & & : + 5788.11 & 0.0070 & 0.0065 & & he ii & 5789.65 & 4.39 & 5g+ 2 g & 39h+ 2h * & 50 & & 20.00 + 5793.58 & 0.0060 & 0.0056 & & he ii & 5794.57 & 5.38 & 5g+ 2 g & 38h+ 2h * & 50 & & 36.67 + 5796.19 & 0.0050 & 0.0046 & & fe ii & 5797.29 & & e4f & 4f 2@xmath402@xmath41 * & 8 & 6 & 40.00 + 5799.41 & 0.0070 & 0.0065 & & he ii & 5800.48 & 5.37 & 5g+ 2 g & 37h+ 2h * & 50 & & 2.86 + 5799.83 & 0.0590 & 0.0547 & & c iv & 5801.33 & v1 & 3s 2s & 3p 2p * & 2 & 4 & 3.22 + 5805.51 & 0.0090 & 0.0083 & & he ii & 5806.57 & 5.36 & 5g+ 2 g & 36h+ 2h * & 50 & & 16.67 + 5810.50 & 0.0410 & 0.0380 & & c iv & 5811.98 & v1 & 3s 2s & 3p 2p * & 2 & 2 & 7.32 + 5812.12 & 0.0110 & 0.0102 & & he ii & 5813.19 & 5.35 & 5g+ 2 g & 35h+ 2h * & 50 & & 67.27 + 5819.12 & 0.0150 & 0.0139 & & [ ni iv ] & 5820.10 & & 3d7 4f * & 3d7 4p & 8 & 4 & 22.67 + 5819.42 & 0.0120 & 0.0111 & & he ii & 5820.43 & 5.34 & 5g+ 2 g & 34h+ 2h * & 50 & & 65.83 + 5826.63 & 0.0100 & 0.0093 & & he ii & 5828.60 & 5.33 & 5g+ 2 g & 33h+ 2h * & 50 & & 18.00 + & & & * & c ii & 5827.85 & v22 & 3p 4d & 3d 4p * & 4 & 4 & + 5829.41 & 0.0040 & 0.0037 & & fe iii & 5830.61 & & 2i 5s3i & 2i 5p3h * & 11 & 9 & 62.50 + & & & * & ni ii & 5830.76 & & 3f 6s2f & 2psp2d * & 6 & 4 & + 5831.75 & 0.0050 & 0.0046 & & fe ii & 5832.89 & & 3f 4p4 g * & 3f 4d4f & 6 & 6 & 56.00 + 5835.69 & 0.0140 & 0.0129 & & he ii & 5836.50 & 5.32 & 5g+ 2 g & 32h+ 2h * & 50 & & 8.57 + 5838.97 & 0.0030 & 0.0028 & & ni ii & 5840.08 & & 3p 4d4p & 2f sp4d * & 6 & 8 & 36.67 + 5842.39 & 0.0040 & 0.0037 & & c ii & 5843.62 & v22 & 3p 4d & 3d 4p * & 6 & 6 & 30.00 + 5845.67 & 0.0180 & 0.0166 & & he ii & 5847.10 & 5.31 & 5g+ 2 g & 31h+ 2h * & 50 & & 7.22 + 5855.86 & 0.0170 & 0.0157 & & he ii & 5857.27 & 5.30 & 5g+ 2 g & 30h+ 2h * & 50 & & : + 5860.17 & 0.0410 & 0.0379 & & [ mn v ] & 5861.00 & & 3d3 4f & 3d3 2 g & 8 & 8 & : + 5867.63 & 0.0140 & 0.0129 & & he ii & 5869.00 & 5.29 & 5g+ 2 g & 29h+ 2h * & 50 & & 0.71 + & & & * & ni ii & 5868.89 & & 4s2 4f & 3p 4p4d * & 4 & 6 & + 5874.21 & 15.6290 & 14.4191 & & he i & 5875.60 & v11 & 2p 3p * & 3d 3d & 9 & 15 & 0.97 + 5880.73 & 0.0160 & 0.0148 & & he ii & 5882.14 & 5.28 & 5g+ 2 g & 28h+ 2h * & 50 & & 1.25 + 5884.80 & 0.0290 & 0.0267 & & c ii & 5886.05 & & 2p3 2p * & 7s 2s & 4 & 2 & : + & & & * & [ mn v ] & 5885.40 & & 3d3 4f & 3d3 2 g & 10 & 10 & + 5904.75 & 0.0140 & 0.0129 & & [ ni iv ] & 5905.50 & & 3d7 4f & 3d7 4p & 8 & 6 & 10.71 + 5911.69 & 0.0170 & 0.0156 & & he ii & 5913.26 & 5.26 & 5g+ 2 g & 26h+ 2h * & 50 & & 9.41 + & & & * & c i & 5912.58 & v71 & 3p 1d & 9d 1f * & 5 & 7 & + 5926.42 & 0.0080 & 0.0074 & & n ii & 5927.81 & v28 & 3p 3p & 3d 3d * & 1 & 3 & 12.50 + 5930.56 & 0.0156 & 0.0144 & & he ii & 5931.83 & 5.25 & 5g+ 2 g & 25h+ 2h * & 50 & & : + & 0.0134 & 0.0120 & & n ii & 5931.78 & v28 & 3p 3p & 3d 3d * & 3 & 5 & : + 5940.26 & 0.0323 & 0.0296 & & n ii & 5941.65 & v28 & 3p 3p & 3d 3d * & 5 & 7 & 9.58 + & & & * & n ii & 5940.24 & v28 & 3p 3p & 3d 3d * & 3 & 3 & + 5943.96 & 0.0020 & 0.0018 & & n ii & 5945.37 & & 4d 3f * & 6f 1@xmath404@xmath41 & 5 & 7 & 50.00 + 5951.66 & 0.0240 & 0.0220 & & he ii & 5952.93 & 5.24 & 5g+ 2 g & 24h+ 2h * & 50 & & 4.58 + & & & * & n ii & 5952.39 & v28 & 3p 3p & 3d 3d * & 5 & 5 & + 5956.11 & 0.0030 & 0.0028 & & ar ii & 5957.46 & & 4d 2d & 4f2@xmath404@xmath41 * & 6 & 8 & 30.00 + 5965.03 & 0.0050 & 0.0046 & & ni ii & 5966.33 & & 4f 3@xmath402@xmath41 * & 3f
7d2f & 6 & 6 & 20.00 + 5968.84 & 0.0080 & 0.0073 & & v ii & 5969.93 & v126 & b3d & z5d & 7 & 5 & 12.50 + 5975.76 & 0.0320 & 0.0293 & & he ii & 5977.03 & 5.23 & 5g+ 2 g & 23h+ 2h * & 50 & & 3.75 + 5985.65 & 0.0040 & 0.0037 & & fe iii & 5986.66 & & 3d & 3d * & 7 & 7 & 27.50 + 5989.23 & 0.0150 & 0.0137 & & [ mn v ] & 5990.10 & & 3d3 4f & 3d3 4p & 6 & 6 & 7.33 + 6003.42 & 0.0310 & 0.0284 & & he ii & 6004.72 & 5.22 & 5g+ 2 g & 22h+ 2h * & 50 & & 5.16 + & & & * & fe ii & 6004.69 & & 4s2 4f & 3f 4p4 g * & 10 & 10 & + & & & * & ni ii & 6004.72 & & 4f 3@xmath402@xmath41 * & 3f 7d4f & 6 & 4 & + 6023.09 & 0.0120 & 0.0110 & & [ mn v ] & 6024.40 & & 3d3 4f & 3d3 4p & 4 & 4 & 10.83 + 6035.39 & 0.0330 & 0.0301 & & he ii & 6036.78 & 5.21 & 5g+ 2 g & 21h+ 2h * & 50 & & 4.85 + & & & * & fe iii & 6036.55 & & 4 g 5s5 g & 4 g 5p5 g * & 13 & 13 & + 6042.29 & 0.0030 & 0.0027 & & ni ii & 6043.38 & & 4fsp4 g * & 5 g 4@xmath405@xmath41 & 12 & 12 & 43.33 + 6047.97 & 0.0030 & 0.0027 & & ne ii & 6049.02 & & 4d 2d & 6f 1@xmath403@xmath41 * & 6 & 8 & 43.33 + 6064.68 & 0.0090 & 0.0082 & & n ii & 6065.92 & & 4d 3d * & 6f 2@xmath402@xmath41 & 3 & 5 & 14.44 + 6072.79 & 0.0380 & 0.0346 & & he ii & 6074.10 & 5.20 & 5g+ 2 g & 20h+ 2h * & 50 & & 4.47 + & & & * & fe iii & 6081.87 & & 4d 5p3f & 4 g 5d5f & 9 & 11 & + 6082.42 & 0.0250 & 0.0228 & & [ mn v ] & 6083.20 & & 3d3 4f & 3d3 4p & 4 & 2 & 6.00 + 6094.60 & 0.1660 & 0.1509 & & [ fe iii ] & 6096.30 & & 3d6 3p2 & 3d6 1d2 & 5 & 5 & 0.78 + 6100.38 & 0.0090 & 0.0082 & & [ k iv ] & 6101.83 & f1 & 3p4 3p & 3p4 1d & 5 & 5 & 26.67 + & & & * & ne ii & 6101.42 & & 4d 4d & 6f 2@xmath405@xmath41 * & 8 & 10 & + 6105.79 & 0.0090 & 0.0082 & & [ kr iv ] & 6107.80 & & 5p3 2d & 5d3 2p & 4 & 4 & 15.56 + 6109.97 & 0.0060 & 0.0055 & & fe i & 6111.02 & & y5go & 9/2[13/2 ] & 13 & 13 & 23.33 + 6113.38 & 0.0090 & 0.0082 & & ni i & 6114.47 & & 3p sp5p * & f3p & 7 & 5 & 17.78 + & & & * & n ii & 6114.60 & v36 & 3d 3f * & 4p 3d & 5 & 7 & + 6116.98 & 0.0440 & 0.0400 & & he ii & 6118.26 & 5.19 & 5g+ 2 g & 19h+ 2h * & 50 & & 3.18 + 6123.69 & 0.0100 & 0.0091 & & [ ni iv ] & 6124.10 & & 3d7 4f & 3d7 4p & 6 & 4 & 13.00 + 6128.52 & 0.0050 & 0.0045 & & fe ii & 6129.69 & v46 & a4 g & z6f * & 10 & 10 & 26.00 + 6130.98 & 0.0040 & 0.0036 & & [ cr iii ] & 6132.54 & & 3d4 3 g & 3d4 1f & 7 & 7 & 27.50 + 6135.08 & 0.0050 & 0.0045 & & n ii & 6136.89 & v36 & 3d 3f * & 4p 3d & 7 & 7 & 14.00 + 6139.21 & 0.0070 & 0.0063 & & ne ii & 6140.44 & & 4f 2@xmath404@xmath41 * & 6 g 0@xmath404@xmath41 & 8 & 10 & 34.29 + 6142.66 & 0.0060 & 0.0054 & & si i & 6143.93 & & 3s & ( 1/2,3/2 ) * & 3 & 3 & 50.00 + 6149.46 & 0.0006 & 0.0005 & & n ii & 6150.75 & v36 & 3d 3f * & 4p 3d & 5 & 5 & 10.00 + 6149.86 & 0.0306 & 0.0277 & & c ii & 6151.43 & v16.04 & 4d 2d & 6f 2f * & 16 & 20 & 11.64 + 6154.68 & 0.0040 & 0.0036 & & o i & 6155.97 & v10 & 3p 5p & 4d 5d * & 3 & 5 & 27.50 + 6155.47 & 0.0070 & 0.0063 & & o i & 6156.77 & v10 & 3p 5p & 4d 5d * & 5 & 7 & 27.14 + 6156.04 & 0.0140 & 0.0127 & & [ mn v ] & 6157.60 & & 3d3 4f & 3d3 4p & 6 & 4 & 69.29 + 6156.88 & 0.0090 & 0.0082 & & o i & 6158.17 & v10 & 3p 5p & 4d 5d * & 7 & 9 & 28.89 + 6160.54 & 0.0070 & 0.0063 & & [ cl ii ] & 6161.84 & & 3p4 1d & 3p4 1s & 5 & 1 & 32.86 + 6164.70 & 0.0210 & 0.0190 & & [ mn v ] & 6166.00 & & 3d3 4f & 3d3 4p & 8 & 6 & 12.86 + 6166.46 & 0.0080 & 0.0072 & & n ii & 6167.76 & v36 & 3d 3f * & 4p 3d & 9 & 7 & 25.00 + 6168.87 & 0.0030 & 0.0027 & & n ii & 6170.16 & v36 & 3d 3f * & 4p 3d & 5 & 3 & 20.00 + 6169.30 & 0.0480 & 0.0434 & & he ii & 6170.60 & 5.18 & 5g+ 2 g & 18h+ 2h * & 50 & & 14.58 + 6172.01 & 0.0050 & 0.0045 & & n ii & 6173.31 & v36 & 3d 3f * & 4p 3d & 7 & 5 & 26.00 + 6178.31 & 0.0070 & 0.0063 & & [ co iii ] & 6179.60 & & a4d & b4p & 2 & 2 & 30.00 + 6184.93 & 0.0100 & 0.0090 & & fe iii & 6186.08 & & 4p 5s5p & 4p 5p5d * & 5 & 5 & : + 6187.60 & 0.0080 & 0.0072 & & ne ii & 6188.72 & & 4d 4d & 6f 2@xmath401@xmath41 * & 4 & 6 & : + 6189.51 & 0.0120 & 0.0108 & & o i & 6191.00 & & 3d * & 3d & 3 & 7 & : + 6192.44 & 0.0170 & 0.0154 & & [ fe iv ] & 6193.90 & & 4p & 2f3 & 6 & 8 & 87.65 + 6195.76 & 0.0260 & 0.0235 & & [ v ii ] & 6197.12 & & a5d & b3 g & 7 & 7 & 21.15 + 6199.39 & 0.0160 & 0.0145 & & si i & 6200.46 & & 3p & ( 3/2,1/2 ) * & 1 & 3 & 15.62 + 6203.98 & 0.0190 & 0.0172 & & [ fe iii ] & 6205.35 & & 3p2 & 5p & 5 & 5 & 8.95 + 6212.30 & 0.0080 & 0.0072 & & o i & 6213.59 & & 3d * & 3d & 7 & 7 & 11.25 + 6216.96 & 0.0100 & 0.0090 & & o i & 6218.16 & & 3d * & 3d & 7 & 5 & 4.00 + 6218.11 & 0.0140 & 0.0126 & & [ mn v ] & 6219.10 & & 3d3 2d2 & 3d3 2f & 6 & 2 & 3.57 + 6222.40 & 0.0090 & 0.0081 & & fe iii & 6223.47 & & 6s 5f5f * & 4 g 6s5 g & 7 & 9 & 10.00 + 6227.26 & 0.0060 & 0.0054 & & [ k vi ] & 6228.60 & & 3p2 3p & 3p2 1d & 5 & 5 & 15.00 + & & & * & fe ii ] & 6229.35 & v34 & b4f & z6d & 8 & 6 & + 6232.46 & 0.0420 & 0.0379 & & fe ii & 6233.75 & & 5d 4d4p & 4f 0@xmath403@xmath41 * & 6 & 8 & 3.10 + 6232.66 & 0.0160 & 0.0144 & & he ii & 6233.82 & 5.17 & 5g+ 2 g & 17h+ 2h * & 50 & & 3.75 + & & & * & [ mn v ] & 6234.00 & & 3d3 2d2 & 3d3 2f & 4 & 6 & + 6237.28 & 0.0050 & 0.0045 & & fe ii & 6238.39 & v74 & b4d & z4p & 4 & 4 & 40.00 + 6240.06 & 0.0060 & 0.0054 & & ne ii & 6241.18 & & 4d 4p & 6f 1@xmath403@xmath41 * & 4 & 6 &
28.33 + 6243.19 & 0.0030 & 0.0027 & & [ cr iii ] & 6244.67 & & 3d4 3 g & 3d4 1f & 11 & 7 & 50.00 + 6249.14 & 0.0130 & 0.0117 & & [ cr iii ] & 6250.38 & & 3d4 b1 g & 3d4 b1d & 9 & 5 & 10.00 + 6253.26 & 0.0070 & 0.0063 & & [ fe ii ] & 6254.30 & & a6d & a2 g & 8 & 8 & 20.00 + 6257.14 & 0.0130 & 0.0117 & & [ fe v ] & 6258.36 & & 3d4 3d & 3d4 1f & 5 & 7 & 10.77 + 6262.15 & 0.0060 & 0.0054 & & fe i & 6263.22 & v863 & b3d & x3d * & 3 & 5 & 21.67 + 6265.63 & 0.0070 & 0.0063 & & o i & 6266.89 & v48 & 3p 3f & 4d 3f * & 9 & 9 & 20.00 + 6271.67 & 0.0100 & 0.0090 & & ne i & 6273.02 & & 1/2@xmath401/2@xmath41 & 1/2@xmath401/2@xmath41 * & 3 & 3 & 10.00 + 6281.47 & 11.1110 & 9.9853 & & dib & 6284.31 & & & & & & + 6286.75 & 0.0020 & 0.0018 & & n ii & 6288.18 & & 3p & 3d * & 5 & 3 & : + 6298.82 & 0.6230 & 0.5593 & & [ o i ] & 6300.30 & f1 & 2p4 3p & 2p4 1d & 5 & 5 & 0.74 + 6309.56 & 0.1000 & 0.0897 & & he ii & 6310.80 & 5.16 & 5g+ 2 g & 16h+ 2h * & 50 & & 0.40 + 6310.78 & 1.5490 & 1.3895 & & [ s iii ] & 6312.10 & f3 & 2p2 1d & 2p2 1s & 5 & 1 & 0.37 + 6342.12 & 0.0280 & 0.0251 & & [ mn v ] & 6343.60 & & 3d3 4f & 3d3 4p & 8 & 4 & 13.93 + & & & * & [ ni iv ] & 6343.40 & & 3d7 2h & 3d7 2f & 10 & 6 & + 6345.62 & 0.0550 & 0.0492 & & si ii & 6347.10 & v2 & 4s 2s & 4p 2p * & 2 & 4 & 7.09 + & & & * & mg ii & 6346.80 & v16 & 4d 2d & 6f 2f * & 10 & 14 & + & & & * & ni ii & 6347.03 & & 3f 5d2h & 6f 3@xmath404@xmath41 * & 12 & 10 & + 6362.29 & 0.2160 & 0.1932 & & [ o i ] & 6363.78 & f1 & 2p4 3p & 2p4 1d & 3 & 5 & 1.81 + 6370.10 & 0.0840 & 0.0751 & & si ii & 6371.36 & v2 & 4s 2s & 4p 2p * & 2 & 2 & 4.52 + & & & * & ni ii & 6371.47 & & 4f 2@xmath402@xmath41 * & 3f 7d4f & 4 & 4 & + 6386.54 & 0.0040 & 0.0036 & & ne ii & 6387.77 & & 2f & 0@xmath403@xmath41 * & 6 & 6 & 17.50 + 6392.39 & 0.0490 & 0.0437 & & [ mn v ] & 6393.50 & & 3d3 4f & 3d3 4p & 10 & 6 & 2.04 + 6396.84 & 0.0060 & 0.0054 & & o ii & 6398.06 & & g@xmath403@xmath41 * & 2@xmath404@xmath41 & 8 & 10 & 16.67 + 6400.26 & 0.0070 & 0.0062 & & [ ni iii ] & 6401.40 & & 3d8 3f & 3d8 3p & 7 & 3 & 14.29 + & & & * & n ii & 6401.44 & & 4d 1f * & 6f 2@xmath402@xmath41 & 7 & 5 & + 6405.08 & 0.0780 & 0.0696 & & he ii & 6406.38 & 5.15 & 5g+ 2 g & 15h+ 2h * & 50 & & 1.67 + 6410.08 & 0.0050 & 0.0045 & & n ii & 6411.54 & & 4d 1f * & 6f 2@xmath405@xmath41 & 7 & 9 & 16.00 + 6414.38 & 0.0040 & 0.0036 & & fe ii & 6415.57 & & x4p * & f4d & 6 & 8 & 25.00 + 6417.74 & 0.0050 & 0.0045 & & ni ii & 6419.12 & & 4f 2@xmath403@xmath41 * & 3p 5d4f & 8 & 6 & 20.00 + 6422.02 & 0.0030 & 0.0027 & & si i & 6423.25 & & 3d & ( 1/2,1/2 ) * & 3 & 3 & 50.00 + 6424.76 & 0.0070 & 0.0062 & & o ii & 6425.90 & & g@xmath404@xmath41 * & 3@xmath405@xmath41 & 10 & 12 & 21.43 + 6429.39 & 0.0060 & 0.0053 & & cr ii & 6430.56 & & 5d 4f4h * & 6 g 3@xmath407@xmath41 & 12 & 14 & 15.00 + 6433.46 & 0.0280 & 0.0249 & & [ ar v ] & 6435.10 & & 3p2 3p & 3p2 1d & 3 & 5 & 11.07 + 6435.72 & 0.0040 & 0.0036 & & [ cr v ] & 6436.20 & & 3d2 3f & 3d2 3p & 7 & 5 & 80.00 + 6439.76 & 0.0030 & 0.0027 & & [ ni ii ] & 6441.29 & & 4s 2f & 4s 2p & 8 & 4 & 26.67 + 6444.64 & 0.0080 & 0.0071 & & ne ii & 6445.81 & & 2@xmath402@xmath41 * & 2f & 6 & 8 & 10.00 + 6450.55 & 0.0090 & 0.0080 & & v ii & 6451.97 & & b1f & z1 g * & 7 & 9 & 8.89 + 6455.17 & 0.0130 & 0.0116 & & fe ii & 6456.38 & & b4d & z4p * & 8 & 6 & 6.92 + 6460.51 & 0.0874 & 0.0778 & & c ii & 6461.95 & v17.04 & 4f 2f * & 6 g 2 g & 22 & 26 & 1.72 + 6464.61 & 0.0060 & 0.0053 & & fe ii & 6466.07 & & w4d * & 5d 4d4p & 4 & 6 & 16.67 + 6470.18 & 0.0120 & 0.0107 & & [ mn v ] & 6471.30 & & 3d3 2h & 3d3 2f & 10 & 8 & 6.67 + & & & * & o ii & 6471.42 & & 4f d1 * & 6 g 2(2 ) & 6 & 10 & + 6477.14 & 0.0050 & 0.0044 & & n iii & 6478.72 & v14 & 3p 4p & 3d 4d * & 6 & 6 & 32.00 + 6481.15 & 0.0160 & 0.0142 & & [ fe ii ] & 6482.31 & & ( 3p2)4s 4p & ( 3d)4s 2d & 6 & 6 & 10.62 + & & & * & n ii & 6482.05 & v8 & 3s 1p * & 3p 1p & 3 & 3 & + 6484.95 & 0.0260 & 0.0231 & & si i & 6486.14 & & 1d & ( 3/2,5/2 ) * & 5 & 7 & 6.54 + 6494.30 & 0.0180 & 0.0160 & & cr ii & 6495.65 & & 5d 4f4 g * & 6 g 3@xmath405@xmath41 & 8 & 10 & 13.33 + & & & * & ni ii & 6495.74 & & 3f 5d4 g & 7f 4@xmath404@xmath41 * & 6 & 8 & + 6497.62 & 0.0210 & 0.0186 & & ne ii & 6498.74 & & 2f & 1@xmath403@xmath41 * & 6 & 6 & 15.24 + 6500.31 & 0.0380 & 0.0337 & & o ii & 6501.41 & & 4f d3 * & 6 g 0(4 ) & 8 & 10 & 10.26 + 6509.09 & 0.0360 & 0.0319 & & fe ii & 6510.28 & & 3f 4p4 g * & 3f 4d4h & 10 & 12 & 4.44 + 6514.92 & 0.0220 & 0.0195 & & n ii & 6516.15 & & 4d 1p * & 6f 2@xmath403@xmath41 & 3 & 5 & 6.82 + 6525.76 & 0.0900 & 0.0797 & & [ n ii ] & 6527.24 & f1 & 2p2 3p & 2p2 1d & 1 & 5 & 2.56 + & & & * & he
ii & 6527.11 & 5.14 & 5g+ 2 g & 14h+ 2h * & 50 & & + 6531.78 & 0.0070 & 0.0062 & & ne ii & 6533.02 & & 1@xmath404@xmath41 * & 1@xmath405@xmath41 & 8 & 10 & 20.00 + 6538.02 & 0.0580 & 0.0513 & & ne ii & 6539.14 & & 4f 2@xmath402@xmath41 * & 6 g 2@xmath403@xmath41 & 6 & 8 & 10.34 + & & & * & o ii & 6539.33 & & 4f f2 * & 6 g 2(3 ) & 6 & 8 & + 6546.65 & 6.3600 & 5.6254 & & [ n ii ] & 6548.04 & f1 & 2p2 3p & 2p2 1d & 3 & 5 & 1.57 + 6558.75 & 3.8190 & 3.3755 & & he ii & 6560.10 & 4.6 & 4f+ 2f * & 6g+ 2 g & 32 & 72 & 7.86 + 6561.45 & 301.3570 & 266.3203 & & h i & 6562.80 & h3 & 2p+ 2p * & 3d+ 2d & 8 & 18 & 0.02 + 6576.70 & 0.5308 & 0.4686 & & c ii & 6578.05 & v2 & 3s 2s & 3p 2p * & 2 & 4 & 65.63 + 6581.53 & 0.2654 & 0.2343 & & c ii & 6582.88 & v2 & 3s 2s & 3p 2p * & 2 & 2 & 65.63 + 6581.97 & 17.7170 & 15.6386 & & [ n ii ] & 6583.46 & f1 & 2p2 3p & 2p2 1d & 5 & 5 & 2.82 + 6604.15 & 0.0220 & 0.0194 & & ni ii@xmath42 $ ] & 6605.47 & & 4s2 4f & 1d 4p2d * & 4 & 6 & 13.64 + 6608.84 & 0.0170 & 0.0150 & & n ii & 6610.57 & v31 & 3p 1d & 3d 1f * & 5 & 7 & 20.59 + 6612.40 & 11.1110 & 9.7904 & & dib & 6613.72 & & & & & & + 6616.42 & 0.0090 & 0.0079 & & si i & 6617.78 & & 4p 3p & 8d ( 23 ) * & 5 & 7 & : + 6619.79 & 0.0070 & 0.0062 & & si i & 6621.20 & & 3d 1d * & 7p ( 22 ) & 5 & 3 & : + 6630.15 & 0.0030 & 0.0026 & & [ fe ii ] & 6631.23 & & a4d & b4f & 8 & 4 & : + 6638.14 & 0.0080 & 0.0070 & & ne ii & 6639.30 & & 2@xmath402@xmath41 * & 4f & 6 & 8 & : + 6639.78 & 0.0090 & 0.0079 & & o ii & 6641.05 & v4 & 3s 2p & 3p 2s * & 2 & 2 & 47.78 + 6654.65 & 0.0110 & 0.0097 & & fe ii ] & 6656.00 & & a2f & z6f * & 8 & 6 & : + 6666.25 & 0.0200 & 0.0176 & & he i & 6667.72 & & 3s 3s & 35p 3p * & 3 & 1 & 72.00 + 6670.87 & 0.0240 & 0.0211 & & [ cr iii ] & 6672.40 & & 3d4 a3p & 3d4 a1d & 3 & 5 & : + 6676.51 & 4.1880 & 3.6768 & & he i & 6678.15 & v46 & 2p 1p * & 3d 1d & 3 & 5 & 7.70 + 6680.28 & 0.2260 & 0.1984 & & he ii & 6683.20 & 5.13 & 5g+ 2 g & 13h+ 2h * & 50 & & : + & & & * & [ ni iii ] & 6682.20 & & 3d8 3f & 3d8 3p & 5 & 1 & + 6700.46 & 0.0060 & 0.0053 & & [ ni xv ] & 6701.70 & & 3p2 3p & 3p2 3p & 1 & 3 & : + 6703.05 & 0.0060 & 0.0053 & & he i & 6704.67 & & 3s 3s & 26p 3p * & 3 & 9 & : + 6708.54 & 0.0170 & 0.0149 & & [ cr v ] & 6709.90 & & 3d2 3f & 3d2 3p & 9 & 5 & 23.53 + 6715.48 & 1.5720 & 1.3771 & & [ s ii ] & 6716.44 & f2 & 2p3 4s * & 2p3 2d * & 4 & 6 & 0.64 + 6720.17 & 0.0160 & 0.0140 & & o ii & 6721.40 & v4 & 3s 2p & 3p 2s * & 4 & 2 & 18.75 + 6729.86 & 2.6090 & 2.2837 & & [ s ii ] & 6730.82 & f2 & 2p3 4s * & 2p3 2d * & 4 & 4 & 0.77 + & & & * & c iii & 6731.04 & v3 & 3s 3p * & 3p 3d & 3 & 5 & + 6738.93 & 0.0180 & 0.0157 & & [ fe iv ] & 6739.80 & & 4 g & 2i & 12 & 12 & 16.67 + 6743.22 & 0.0190 & 0.0166 & & n ii & 6744.74 & & ?
& ? & & & 15.79 + 6746.51 & 0.0080 & 0.0070 & & [ cr iv ] & 6747.50 & & 3d3 4f & 3d3 2 g & 6 & 8 & 25.00 + 6750.74 & 0.0040 & 0.0035 & & o ii & 6751.86 & & 1@xmath404@xmath41 & h@xmath405@xmath41 * & 8 & 10 & 50.00 + 6754.86 & 0.0070 & 0.0061 & & he i & 6755.85 & & 3s 3s & 20p 3p * & 3 & 9 & 42.86 + 6760.45 & 0.0090 & 0.0079 & & [ fe iv ] & 6761.30 & & 4 g & 2i & 10 & 12 & 22.22 + 6768.16 & 0.0040 & 0.0035 & & he i & 6769.55 & & 3s 3s & 19p 3p * & 3 & 9 & 50.00 + 6779.04 & 0.0080 & 0.0070 & & c ii & 6780.61 & v14 & 3s 4p * & 3p 4d & 2 & 4 & 25.00 + 6785.74 & 0.0060 & 0.0052 & & c ii & 6787.22 & v14 & 3s 4p * & 3p 4d & 2 & 2 & 33.33 + & & & * & si i & 6786.94 & & 3d 3f * & 9f 3f & 7 & 9 & + 6791.01 & 0.0070 & 0.0061 & & fe ii & 6792.54 & & 5d 5p6f * & 5d 5d6 g & 6 & 6 & 28.57 + & & & * & fe ii & 6792.67 & & 4s2 2f & t2f * & 6 & 8 & + 6793.96 & 0.0530 & 0.0462 & & [ k iv ] & 6795.10 & f1 & 3p4 3p & 3p4 1d & 3 & 5 & 5.66 + 6799.62 & 0.0040 & 0.0035 & & fe i & 6800.82 & & w5f * & 9/2@xmath407/2@xmath41 & 7 & 7 & 50.00 + 6846.50 & 0.0130 & 0.0113 & & o ii & 6847.86 & & 2@xmath402@xmath41 & d@xmath402@xmath41 * & 6 & 6 & 23.08 + 6854.98 & 0.0080 & 0.0070 & & he i & 6855.91 & & 3s 3s & 15p 39 * & 3 & 9 & 37.50 + 6863.74 & 0.0090 & 0.0078 & & n ii & 6865.16 & & ? & ? & & & 33.33 + 6890.08 & 0.1060 & 0.0920 & & he ii & 6890.90 & 5.12 & 5g+ 2 g & 12h+ 2h * & 50 & & 4.72 + & & & * & si iii & 6891.32 & & 6s 1s & 7p 1p * & 1 & 3 & + 6933.24 & 0.0200 & 0.0173 & & o ii & 6934.63 & & 3d 2f & 5f d3 * & 8 & 8 & 25.00 + & & & * & fe i & 6934.46 & & y3f * & f5f & 9 & 11 & + 6989.08 & 0.0220 & 0.0190 & & n ii & 6990.70 & & ? & ? & & & 22.73 + 7004.58 & 0.0480 & 0.0414 & & [ ar v ] & 7005.67 & & 3p2 3p & 3p2 1d & 5 & 5 & 10.42 + & & & * & ni ii & 7006.19 & & 3p 4d4f & 5f 3@xmath404@xmath41 * & 6 & 8 & + 7047.15 & 0.0240 & 0.0207 & & [ fe ii ] & 7047.99 & & ( 5d)4s 4d & ( 3f2)4s 4f & 2 & 4 & 16.67 + 7064.17 & 5.0700 & 4.3592 & & he i & 7065.71 & v10 & 2p 3p * & 3s 3s & 9 & 3 & 3.94 + 7101.73 & 0.0250 & 0.0215 & & [ ni ii ] & 7102.66 & & ( 3f)4s 2f & ( 3p)4s 2p & 6 & 4 & 20.00 + 7109.86 & 0.0310 & 0.0266 & & [ fe iv ] & 7111.10 & & 3d5 4d & 3d5 4f & 8 & 6 & 16.13 + 7113.32 & 0.0400 & 0.0343 & & fe i & 7114.55 & v267 & a3 g & z3 g * & 11 & 9 & 12.50 + 7134.75 & 17.4640 & 14.9619 & & [ ar iii ] & 7135.80 & f1 & 3p4 3p & 3p4 1d & 5 & 5 & 1.89 + 7154.26 & 0.0200 & 0.0171 & & [ fe ii ] & 7155.16 & & 3d7 4f & 3d7 2 g & 10 & 10 & 10.00 + 7159.66 & 0.0460 & 0.0394 & & he i & 7160.61 & 3.10 & 3s 3s & 10p 3p & 3 & 9 & 8.70 + 7169.63 & 0.1250 & 0.1069 & & [ fe iv ] & 7170.90 & & 4d & 4f & 2 & 6 & 8.00 + & & & * & [ ar iv ] & 7170.62 & f2 & 3p3 2d * & 3p3 2p * & 4 & 4 & + 7176.56 & 0.1560 & 0.1334 & & he ii & 7177.50 & 5.11 & 5g+ 2 g & 11h+ 2h * & 50 & * & 6.41 + 7189.63 & 0.0350 & 0.0299 & & fe ii & 7191.30 & v72 & b4d & z4f * & 6 & 4 & 22.86 + & & & * & [ fe iv ] & 7190.82 & & 3d5 4d & 3d5 4f & 6 & 6 & + 7196.96 & 0.0280 & 0.0239 & & ca i & 7198.17 & & 3s & 3p * & 3 & 3 & 17.86 + 7213.03 & 0.0250 & 0.0213 & & n ii & 7214.71 & v52 & 3d 3p * & 4p 3d & 5 & 3 & 20.00 + 7220.30 & 0.0240 & 0.0205 & & [ fe ii ] & 7221.84 & & b4p & b2d & 2 & 6 & 20.83 + 7230.26 & 0.1520 & 0.1296 & & c ii & 7231.32 & v3 & 3p 2p * & 3d 2d & 2 & 4 & 3.29 + 7235.98 & 0.3130 & 0.2668 & & [ ar iv ] & 7237.40 & f2 & 3p3 2d * & 3p3 2p * & 6 & 4 & 3.83 + & & & * & c ii & 7236.42 & v3 & 3p 2p * & 3d 2d & 4 & 6 & + & & & * & c ii & 7237.17 & v3 & 3p 2p * & 3d 2d & 4 & 4 & + 7261.89 & 0.1270 & 0.1081 & & [ ar iv ] & 7262.76 & f2 & 3p3 2d * & 3p3 2p * & 4 & 2 & 9.45 + 7280.36 & 0.5940 & 0.5052 & & he i & 7281.35 & v45 & 2p 1p * & 3s 1s & 3 & 1 & 2.53 + 7296.77 & 0.0750 & 0.0637 & & he i & 7298.04 & & 3s 3s & 9p 3p * & 3 & 9 & 13.33 + 7306.65 & 0.0300 & 0.0255 & & fe ii & 7307.97 & v73 & b4d & z4d * & 4 & 4 & 33.33 + 7318.80 & 1.4030 & 1.1911 & & [ o ii ] & 7319.99 & f2 & 2p3 2d * & 2p3 2p * & 6 & 4 & 3.56 + 7329.09 & 1.3080 & 1.1099 & & [ o ii ] & 7330.73 & f2 & 2p3 2d * & 2p3 2p * & 4 & 4 & 3.82 + 7346.44 & 0.0190 & 0.0161 & & [ v ii ] & 7347.75 & & b3d & b1f & 7 & 7 & 52.63 + 7352.30 & 0.0230 & 0.0195 & & fe ii & 7353.92 & & 4s2 2 g & u2 g * & 10 & 10 & 43.48 + 7364.13 & 0.0190 & 0.0161 & & ni ii & 7365.71 & & 1 g 4d2f & 4f"2@xmath401@xmath41 * & 6 & 4 & 52.63 + 7454.07 & 0.0270 & 0.0228 & & si i & 7455.36 & v57.02 & 4p 3d & 5d 3d * & 5 & 3 & 37.04 + 7462.74 & 0.0290 & 0.0245 & & [ cr ii ] & 7463.98 & & a4d & b4 g & 8 & 8 & 34.48 + 7470.66 & 0.0300 & 0.0253 & & ne i & 7472.44 & v8.02 & 3p 2@xmath401@xmath41 & 3d 2@xmath402@xmath41 * & 3 & 3 & 33.33 + 7484.78 & 0.0250 & 0.0211 & & o ii & 7486.21 & & 3d 2 g & 5f f2 * & 8 & 6 & 40.00 + & & & * & o ii & 7486.35 & & 3d 4p & 4p 4d * & 6 & 4 & + 7498.76 & 0.0510 & 0.0429 & & he i & 7499.85 & & 3s 3s & 8p 3p * & 3 & 9 & 19.61 + 7514.29 & 0.0330 & 0.0278 & & fe ii & 7515.83 & v73 & b4d & z4d * & 8 & 6 & 30.30 + 7518.26 & 0.0360 & 0.0303 & & c ii & 7519.86 & v17.13 & 5p 2p * & 3p 2d & 2 & 4 & 27.78 + & & & * & c ii & 7519.50 & v16.08 & 2p3 2p * & 3p 2p & 4 & 4 & + & & & * & c ii & 7519.86 & v16.08 & 2p3 2p * & 3p 2p & 2 & 2 & + 7529.40 & 0.4520 & 0.3800 & & [ cl iv ] & 7530.80 & & 3p2 3p & 3p2 1d & 3 & 5 & 2.21 + 7549.32 & 0.0190 & 0.0160 & & n ii & 7551.00 & & 3p & 3p * & 5 & 3 & 52.63 + 7577.64 & 0.0340 & 0.0285 & & fe ii & 7579.41 & & y4 g * & 4s2 2h & 8 & 10 & 29.41 + 7591.41 & 0.1950 & 0.1635 & & he ii & 7592.75 & 5.10 & 5g+ 2 g & 10h+ 2h * & 50 & & 5.13 + 7674.30 & 0.0300 & 0.0251 & & [ fe ii ] & 7675.30 & & 4s 6d & 3d7 4p & 4 & 2 & 33.33 + 7680.38 & 0.0140 & 0.0117 & & fe ii & 7682.10 & & e4d & 3f 4p4d * & 6 & 6 & 71.43 + 7686.44 & 0.0210 & 0.0175 & & [ fe ii ] & 7687.95 & & ( 5d)4s 6d & 3d7 4p & 6 & 4 & 47.62 + 7705.95 & 0.4070 & 0.3395 & & c iii & 7707.43 & v10.01 & 3p 1p & 3d 1d * & 3 & 5 & 2.46 + 7712.50 & 0.0570 & 0.0475 & & he i & 7714.54 & & 3s 1s & 15p 1p * & 1 & 3 & 17.54 + 7717.52 & 0.1770 & 0.1476 & & si iv & 7718.79 & v19 & 6 g 2 g & 7h 2h * & 10 & 12 & 5.65 + 7723.46 & 0.1880 & 0.1567 & & n ii & 7725.00 & & ? & ? & & & 5.32 + 7727.41 & 0.1710 & 0.1425 & & [ fe ii ] & 7728.66 & & a6s & b2d & 6 & 6 & 5.85 + 7731.27 & 0.1470 & 0.1225 & & [ v ii ] & 7733.19 & & b1 g & d3p & 9 & 5 & 6.80 + 7736.69 & 0.1330 & 0.1108 & & ni ii & 7738.31 & & 4f 4@xmath402@xmath41 * & 3f 6d2f & 4 & 6 & 7.52 + 7749.98 & 4.0400 & 3.3633 & & [ ar iii ] & 7751.10 & & 3p4 3p & 3p4 1d & 3 & 5 & 0.25 + 7762.27 & 0.0700 & 0.0582 & & ni ii & 7763.87 & & 4s2 2f & 3f 5p4f * & 8 & 6 & 10.00 + 7771.91 & 0.0610 & 0.0507 & & ni ii & 7773.62 & & 4f 2@xmath401@xmath41 * & 1d 5d2d & 4 & 4 & 11.48 + 7782.36 & 0.1240 & 0.1031 & & fe ii & 7784.21 & & e4d & v4f * & 8 & 8 & 6.45 + 7803.56 & 0.0830 & 0.0689 & & [ fe ii ] & 7805.01 & & a6s & b2d & 6 & 4 & 9.64 + 7808.82 & 0.0480 & 0.0399 & & cl ii & 7810.40 & & 4p 1p & 3d 1d * & 3 & 5 & 13.54 + 7815.08 & 0.0930 & 0.0772 & & he i & 7816.13 & 3.7 & 3s 3s & 7p 3p * & 3 & 9 & 8.60 + 7831.94 & 0.0680 & 0.0564 & & ni ii & 7833.68 & & 2i & 4@xmath407@xmath41 * & 14 & 14 & 11.76 + 7837.02 & 0.1160 & 0.0962 & & ne ii & 7838.55 & & 3d 2p & 4p 2p * & 2 & 4 & 6.90 + 7858.49 & 0.0700 & 0.0580 & & o ii & 7859.76 & & 3d 2d & 5f d2 * & 4 & 4 & 11.43 + 7861.36 & 0.0580 & 0.0481 & & fe ii & 7862.95 & & 4d sp6f * & 5d 5d6 g & 8 & 10 & 13.79 + 7864.98 & 0.0550 & 0.0456 & & ni ii & 7866.76 , & & 3p 4d2f & 5f 4@xmath404@xmath41 * & 8 & 10 & 14.55 + 7868.54 & 0.0300 & 0.0249 & & [ v ii ] & 7870.00 & & a1 g & d3f & 9 & 9 & 26.67 + 7873.05 & 0.0750 & 0.0621 & & [ fe ii ] & 7874.65 & & ( 5d)4s 6d & 3d7 4p & 2 & 4 & 10.67 + 7877.12 & 0.0330 & 0.0273 & & fe i & 7878.66 & & c3f & y3 g * & 9 & 7 & 24.24 + 7882.12 & 0.0730 & 0.0604 & & n i & 7883.40 & & 2p * & 2p & 4 & 4 & 10.96 + 7888.55 & 0.0870 & 0.0720 & & [ ni iii ] & 7889.90 & & 3d8 3f & 3d8 1d & 7 & 5 & 9.20 + 7895.29 & 0.0410 & 0.0339 & & he i & 7896.89 & & 3p 3p * & 35s 3s & 1 & 3 & 19.51 + 7924.87 & 0.0290 & 0.0240 & & he i & 7926.38 & & 3p 3p * & 29s 3s & 9 & 3 & 27.59 + 7979.14 & 0.0170 & 0.0140 & & n ii & 7980.70 & & 4p 3p & 5s 3p * & 5 & 3 & 47.06 + 8000.21 & 0.0370 & 0.0305 & & ar i & 8001.86 & & 3d 2@xmath401@xmath41 * & 8p 2@xmath401@xmath41 & 3 & 3 & 21.62 + 8012.60 & 0.0380 & 0.0313 & & n iii & 8014.00 & & 3p 2d & 3d 2d * & 6 & 4 & 18.42 + 8016.71 & 0.0370 & 0.0305 & & fe ii & 8018.08 & & 3p 5s4p & 3h 5p4 g * & 6 & 8 & 18.92 + 8020.86 & 0.0260 & 0.0214 & & [ fe ii ] & 8022.65 & & a4d & a2d2 & 2 & 4 & 26.92 + 8032.15 & 0.0330 & 0.0272 & & [ ni ii ] & 8033.54 & & ( 3f)4s 4f & ( 3p)4s 4p & 4 & 6 & 21.21 + 8044.11 & 0.7740 & 0.6366 & & [ cl iv ] & 8045.63 & & 3p2 3p & 3p2 1d & 5 & 5 & 1.29 + 8057.31 & 0.0350 & 0.0288 & & fe ii & 8058.95 & & 4d sp6f * & 5d 5d6d & 10 & 8 & 17.14 + 8066.98 & 0.0170 & 0.0140 & & c iii & 8068.40 & & 5 g 3 g & 6h 3h * & 33 & 27 & 35.29 + 8075.16 & 0.0190 & 0.0156 & & [ v ii ] & 8076.62 & & a3 g & a1f & 7 & 7 & 31.58 + 8081.51 & 0.0310 & 0.0255 & & fe ii & 8083.88 & & z4d * & d4p & 4 & 6 & 19.35 + 8094.43 & 0.0300 & 0.0246 & & ti iii & 8096.26 & & 3 g * & 3/2@xmath409/2@xmath41 & 11 & 9 & 20.00 + 8099.83 & 0.0220 & 0.0181 & & ti iii & 8101.56 & & 3d5f 1f * & 6 g 2@xmath404@xmath41 & 7 & 7 & 27.27 + & & & * & fe iii & 8101.31 & & b2d 4s3d & 4p 4p5p * & 5 & 7 & + 8109.21 & 0.0320 & 0.0262 & & n ii & 8111.00 & & ? & ? & & & 18.75 + 8115.33 & 0.0120 & 0.0098 & & he i & 8116.60 & 3.16 & 3p 3p * & 16d 3d & 9 & & 33.33 + 8121.51 & 0.3480 & 0.2853 & & al ii & 8123.42 & v110 & 5d 3d & 8f 3f * & 5 & 5 & 7.18 + 8131.83 & 0.0250 & 0.0205 & & si i & 8133.90 & & 3d 3f * & 7p ( 11 ) & 5 & 3 & 20.00 + 8139.67 & 0.0290 & 0.0238 & & fe iii & 8141.18 & & 4d 4d5 g & 4 g 5p5h * & 9 & 7 & 17.24 + 8158.27 & 0.0270 & 0.0221 & & fe ii ] & 8159.95 & & f4f & 3f 5p4 g * & 8 & 10 & 18.52 + 8167.51 & 0.0120 & 0.0098 & & he i & 8168.91 & & 3p 3p * & 15s 3s & 9 & 3 & 41.67 + 8173.05 & 0.0210 & 0.0172 & & na i & 8174.86 & & 2d & 2p * & 6 & 6 & 23.81 + 8174.47 & 0.0210 & 0.0172 & & na i & 8176.22 & & 2d & 2p * & 6 & 6 & 23.81 + 8183.57 & 0.0340 & 0.0278 & & ne i & 8185.51 & & 4s 2@xmath402@xmath41 * & 7p 2@xmath401@xmath41 & 3 & 3 & 14.71 + 8184.53 & 0.0160 & 0.0131 & & na i & 8186.56 & & 2d & 2p * & 6 & 6 & 31.25 + 8194.97 & 0.1160 & 0.0948 & & fe i & 8196.61 & & d3f & w3f * & 9 & 7 & 8.62 + 8202.80 & 0.0160 & 0.0131 & & fe ii & 8204.42 & & 4p sp4p * & 3f 4d4p & 2 & 4 & 31.25 + 8214.12 & 0.0180 & 0.0147 & & n ii & 8215.90 & & ? & ? & & & 22.22 + 8220.41 & 0.0480 & 0.0392 & &
[ fe ii ] & 8222.04 & & a4 g & c2d & 6 & 4 & 20.83 + 8223.31 & 0.0430 & 0.0351 & & n ii & 8225.10 & & 3p * & 3p & 5 & 3 & : + 8235.14 & 0.4110 & 0.3355 & & he ii & 8236.77 & 5.9 & 5g+ 2 g & 9h+ 2h * & 50 & & 10.95 + & & & * & he i & 8237.15 & & 3d 3d & 35p 3p * & 5 & 6 & + 8242.11 & 0.0090 & 0.0073 & & h i & 8243.69 & p41 & 3d+ 2d & 41f+ 2f * & 18 & & : + 8249.41 & 0.0140 & 0.0114 & & h i & 8249.97 & p40 & 3d+ 2d & 40f+ 2f * & 18 & & 71.43 + 8253.69 & 0.0040 & 0.0033 & & h i & 8255.02 & p38 & 3d+ 2d & 38f+ 2f * & 18 & & : + 8260.54 & 0.0080 & 0.0065 & & h i & 8260.93 & p36 & 3d+ 2d & 36f+ 2f * & 18 & & : + 8263.12 & 0.0260 & 0.0212 & & h i & 8264.28 & p35 & 3d+ 2d & 35f+ 2f * & 18 & & 38.46 + & & & * & he i & 8265.71 & 3.9 & 3s 1s & 9p 1p * & 1 & 3 & + 8266.63 & 0.0220 & 0.0179 & & h i & 8267.94 & p34 & 3d+ 2d & 34f+ 2f * & 18 & & 45.45 + 8270.24 & 0.0130 & 0.0106 & & h i & 8271.93 & p33 & 3d+ 2d & 33f+ 2f * & 18 & & 76.92 + 8275.28 & 0.0200 & 0.0163 & & h i & 8276.31 & p32 & 3d+ 2d & 32f+ 2f * & 18 & & 50.00 + 8279.77 & 0.0240 & 0.0196 & & h i & 8281.12 & p31 & 3d+ 2d & 31f+ 2f * & 18 & & 48.75 + 8284.91 & 0.0460 & 0.0375 & & h i & 8286.43 & p30 & 3d+ 2d & 30f+ 2f * & 18 & & 25.43 + 8290.70 & 0.0560 & 0.0456 & & h i & 8292.31 & p29 & 3d+ 2d & 29f+ 2f * & 18 & & 20.89 + 8297.13 & 0.0790 & 0.0643 & & h i & 8298.83 & p28 & 3d+ 2d & 28f+ 2f * & 18 & & 14.81 + 8304.51 & 0.0830 & 0.0676 & & h i & 8306.11 & p27 & 3d+ 2d & 27f+ 2f * & 18 & & 13.98 + 8312.42 & 0.1120 & 0.0912 & & h i & 8314.26 & p26 & 3d+ 2d & 26f+ 2f * & 18 & & 10.45 + 8321.59 & 0.1160 & 0.0944 & & h i & 8323.42 & p25 & 3d+ 2d & 25f+ 2f * & 18 & & 10.00 + 8331.94 & 0.1510 & 0.1228 & & h i & 8333.78 & p24 & 3d+ 2d & 24f+ 2f * & 18 & & 7.81 + 8343.53 & 0.1960 & 0.1594 & & h i & 8345.47 & p23 & 3d+ 2d & 23f+ 2f * & 18 & & 6.12 + 8357.64 & 0.2810 & 0.2283 & & h i & 8359.00 & p22 & 3d+ 2d & 22f+ 2f * & 18 & & 4.45 + 8372.81 & 0.2130 & 0.1730 & & h i & 8374.48 & p21 & 3d+ 2d & 21f+ 2f * & 18 & & 5.73 + 8390.68 & 0.2320 & 0.1883 & & h i & 8392.40 & p20 & 3d+ 2d & 20f+ 2f * & 18 & & 5.34 + 8411.59 & 0.2980 & 0.2417 & & h i & 8413.32 & p19 & 3d+ 2d & 19f+ 2f * & 18 & & 4.36 + 8420.39 & 0.0260 & 0.0211 & & he i & 8421.99 & 3.18 & 3d 3d & 18p 3p * & 15 & & 19.23 + 8436.22 & 0.3000 & 0.2431 & & h i & 8437.95 & p18 & 3d+ 2d & 18f+ 2f * & 18 & & 4.37 + 8443.15 & 0.0260 & 0.0211 & & he i & 8444.69 & & 3p 3p * & 11d 3d & 9 & 15 & 19.23 + 8449.72 & 0.0150 & 0.0121 & & he i & 8451.20 & 3.17 & 3d 3d & 17p 3p * & 15 & & 33.33 + 8465.50 & 0.3910 & 0.3165 & & h i & 8467.25 & p17 & 3d+ 2d & 17f+ 2f * & 18 & & 3.61 + 8479.12 & 0.0250 & 0.0202 & & [ cl iii ] & 8480.85 & & 3p3 2d * & 3p3 2p * & 6 & 4 & 20.00 + 8484.83 & 0.0270 & 0.0218 & & he i & 8486.31 & 3.16 & 3d 3d & 16p 3p * & 15 & & 18.52 + 8500.67 & 0.4270 & 0.3452 & &
h i & 8502.48 & p16 & 3d+ 2d & 16f+ 2f * & 18 & & 3.35 + 8516.92 & 0.0170 & 0.0137 & & he ii & 8519.35 & 6.31 & 6s+ 2s & 31p+ 2p * & 72 & & 23.53 + 8527.62 & 0.0490 & 0.0396 & & he i & 8528.99 & 3.15 & 3d 3d & 15p 3p * & 15 & & 8.16 + 8543.64 & 0.5130 & 0.4141 & & h i & 8545.38 & p15 & 3d+ 2d & 15f+ 2f * & 18 & & 2.81 + 8560.50 & 0.0080 & 0.0065 & & [
v ii ] & 8561.99 & & b3d & d3p & 7 & 3 & 25.00 + 8565.62 & 0.0260 & 0.0210 & & he ii & 8566.92 & 6.29 & 6s+ 2s & 29p+ 2p * & 72 & & 15.38 + 8577.20 & 0.0690 & 0.0556 & & [ cl ii ] & 8578.69 & & 3p4 3p & 3p4 1d & 5 & 5 & 7.25 + 8580.50 & 0.0630 & 0.0508 & & he i & 8581.87 & & 3d 3d & 14f 3f * & 15 & 21 & 7.94 + 8596.61 & 0.6110 & 0.4923 & & h i & 8598.39 & p14 & 3d+ 2d & 14f+ 2f * & 18 & & 2.41 + 8606.63 & 0.0060 & 0.0048 & & he i & 8608.31 & & 3p 1p * & 15d 1d & 3 & 5 & 33.33 + 8615.42 & 0.0200 & 0.0161 & & [ fe ii ] & 8616.95 & & 4s 4h & 4s 2i & 14 & 14 & 15.00 + 8624.53 & 0.0280 & 0.0225 & & he ii & 8626.19 & 6.27 & 6s+ 2s & 27p+ 2p * & 72 & & 10.71 + 8631.36 & 0.0130 & 0.0105 & & he i & 8632.97 & 3.10 & 3p 3p * & 10s 3s & 9 & 9 & 30.77 + 8646.98 & 0.0450 & 0.0362 & & he i & 8648.25 & 3.13 & 3d 3d & 13f 3f * & 15 & & 8.89 + 8663.22 & 0.8980 & 0.7219 & & h i & 8665.02 & p13 & 3d+ 2d & 13f+ 2f * & 18 & & 1.71 + 8674.89 & 0.0160 & 0.0129 & & ne ii & 8676.35 & & 4d 4d & 5f 1@xmath403@xmath41 * & 4 & 6 & 18.75 + 8678.89 & 0.0190 & 0.0153 & &
n i & 8680.28 & v1 & 3s 4p & 3p 4d * & 6 & 8 & 15.79 + & & & * & ca iii & 8680.53 & & 5d 2@xmath404@xmath41 * & 5f 2@xmath403@xmath41 & 9 & 7 & + 8684.21 & 0.0170 & 0.0137 & & n i & 8686.15 & v1 & 3s 4p & 3p 4d * & 2 & 4 & 17.65 + 8689.80 & 0.0050 & 0.0040 & & fe iii & 8691.28 & & 4d 4d3 g & 4 g 5p3 g * & 9 & 9 & 54.00 + 8700.14 & 0.0160 & 0.0128 & & n ii & 8701.83 & v65.02a & 4d 3f * & 5f 2@xmath404@xmath41 & 7 & 7 & 18.75 + 8702.94 & 0.0070 & 0.0056 & & [ ni ii ] & 8703.87 & & ( 3f)4s 2f & ( 1d)4s 2d & 8 & 6 & 28.57 + 8708.64 & 0.0060 & 0.0048 & & n ii & 8710.54 & v65.02b & 4d 3f * & 5f 2@xmath405@xmath41 & 9 & 11 & : + 8727.43 & 0.0230 & 0.0184 & & [ fe iii ] & 8728.90 & & 3p4 & 3d & 5 & 7 & 13.04 + 8731.87 & 0.0940 & 0.0754 & & he i & 8733.43 & 3.12 & 3d 3d & 12f 3f * & 15 & & 10.64 + 8738.06 & 0.0090 & 0.0072 & & he i & 8739.97 & & 3d 3d & 12p 3p * & 5 & 3 & 55.56 + 8748.67 & 0.8890 & 0.7126 & & h i & 8750.47 & p12 & 3d+ 2d & 12f+ 2f * & 18 & & 1.79 + 8774.88 & 0.0470 & 0.0376 & & he i & 8776.97 & 3.9 & 3p 3p * & 9d 3d & 1 & 3 & 10.64 + 8787.29 & 0.0130 & 0.0104 & & o ii & 8788.83 & & 4d 4f & 5f g4 * & 6 & 8 & 30.77 + 8791.99 & 0.0400 & 0.0320 & & c ii & 8793.80 & v28.01 & 3p 2d & 3d 2f * & 6 & 8 & 10.00 + 8793.20 & 0.0061 & 0.0049 & & [ cr ii ] & 8795.17 & & a2d & b2d & 4 & 6 & 18.75 + 8797.20 & 0.0110 & 0.0088 & & he ii & 8799.00 & 6.23 & 6s+ 2s & 23p+ 2p * & 72 & & 18.18 + 8798.10 & 0.0280 & 0.0224 & & c ii & 8799.90 & v28.01 & 3p 2d & 3d 2f * & 4 & 6 & 14.29 + 8807.99 & 0.0180 & 0.0144 & & fe ii & 8809.83 & & b4 g & x4 g * & 12 & 12 & 11.11 + 8814.86 & 0.0110 & 0.0088 & & he i & 8816.64 & & 3p 1p * & 12d 1d & 3 & 5 & 18.18 + 8820.68 & 0.0080 & 0.0064 & & o ii & 8822.18 & & 4d 4f & 5f g3 * & 6 & 8 & 25.00 + & & & * & n ii & 8822.73 & v65.01 & 4d 3f * & 5f 1@xmath404@xmath41 & 7 & 7 & + 8844.14 & 0.0930 & 0.0743 & & he i & 8845.37 & 3.11 & 3d 3d & 11f 3f * & 15 & & 5.38 + 8860.93 & 1.4550 & 1.1620 & & h i & 8862.78 & p11 & 3d+ 2d & 11f+ 2f * & 18 & & 1.23 + 8897.79 & 0.0100 & 0.0080 & & c i & 8899.31 & & 3d 1d * & 8p 1d & 5 & 5 & 20.00 + 8913.02 & 0.0160 & 0.0128 & & he i & 8914.77 & 3.7 & 3s 1s & 7p 1p * & 1 & 3 & 12.50 + 8927.43 & 0.0400 & 0.0319 & & he ii & 8929.11 & 6.21 & 6s+ 2s & 21p+ 2p * & 72 & & 7.50 + 8955.55 & 0.0040 & 0.0032 & & si iv & 8957.25 & v3.01 & 5s 2s & 5p 2p * & 2 & 4 & 50.00 + 8969.09 & 0.0110 & 0.0088 & & fe ii ] & 8970.93 & & a2f & z6d * & 8 & 10 & 18.18 + & & & * & ne ii & 8970.83 & & 4p 4p * & 4d 4d & 2 & 4 & + 8977.90 & 0.0170 & 0.0135 & & ni ii & 8980.69 & & 3f 5s2f & 3f 5p2d * & 6 & 6 & 17.65 + 8982.90 & 0.0090 & 0.0072 & & o ii & 8984.40 & & 2d & p@xmath401@xmath41 * & 6 & 4 & 33.33 + 8995.48 & 0.0820 & 0.0652 & & he i & 8997.02 & 3.10 & 3d 3d & 10f 3f * & 15 & & 6.10 + 9012.94 & 1.8210 & 1.4473 & & h i & 9014.91 & p10 & 3d+ 2d & 10f+ 2f * & 18 & & 1.09 + 9019.60 & 0.0120 & 0.0095 & & o ii & 9021.02 & & 4d 4d & 5f f2 * & 4 & 6 & 25.00 + 9023.37 & 0.0050 & 0.0040 & & o ii & 9026.05 & & 4d 4d & 5f f2 * & 4 & 4 & 60.00 + 9029.30 & 0.0090 & 0.0071 & & fe ii ] & 9031.20 & & 4f 4@xmath404@xmath41 * & 5d 6d6d & 8 & 10 & 27.78 + 9037.09 & 0.0140 & 0.0111 & & c ii & 9039.10 & & 4p * & 4d & 6 & 8 & 17.86 + 9048.52 & 0.0070 & 0.0056 & & o ii & 9050.26 & & 4d 4d & 5f d1 * & 2 & 2 & 35.71 + 9066.91 & 29.1710 & 23.1453 & & [ s iii ] & 9068.60 & & 3p2 3p & 3p2 1d & 3 & 5 & 0.51 + 9082.89 & 0.0120 & 0.0095 & & ne ii & 9084.60 & & 3d 4f & 4p 4d * & 6 & 4 & 25.00 + 9093.62 & 0.0190 & 0.0151 & & c i & 9095.46 & & 3d 3d * & 9p 3d & 5 & 7 & 15.79 + & & & * & ne ii & 9095.94 & & 4p 4d * & 4d 4f & 8 & 10 & + 9103.04 & 0.0170 & 0.0135 & & o ii & 9104.95 & & 4d 4d & 5f f4 * & 8 & 10 & 14.71 + 9106.63 & 0.0220 & 0.0174 & & he ii & 9108.54 & 6.19 & 6s+ 2s & 19p+ 2p * & 72 & & 11.36 + 9113.11 & 0.0110 & 0.0087 & & ne ii & 9114.78 & & 4d 4f & 5f 0@xmath403@xmath41 * & 8 & 6 & 22.73 + 9121.78 & 0.0270 & 0.0214 & & [ cl ii ] & 9123.60 & & 3p4 3p & 3p4 1d & 3 & 5 & 11.11 + 9133.49 & 0.0020 & 0.0016 & & o ii & 9135.08 & & 4d 4d & 5f f2 * & 8 & 6 & : + 9161.00 & 0.0190 & 0.0150 & & [ fe iii ] & 9162.53 & & 7s & 5s & 7 & 5 & 13.16 + 9171.81 & 0.0170 & 0.0134 & & he i & 9174.49 & 3.8 & 3p 3p * & 8s 3s & 9 & 3 & 14.71 + 9180.68 & 0.0160 & 0.0127 & & c i & 9182.83 & v48 & 3p 1d & 4d 1d * & 5 & 5 & 15.62 + 9189.26 & 0.0100 & 0.0079 & & ne ii & 9190.96 & & 4p 2f * & 4d 2 g & 8 & 10 & 24.00 + 9198.25 & 0.0070 & 0.0055 & & ne ii & 9199.74 & & 4p 4s * & 5s 4p & 4 & 4 & 34.29 + 9208.66 & 0.1370 & 0.1082 & & he i & 9210.33 & v83 & 3d 3d & 9f 3f * & 15 & 21 & 7.30 + 9226.97 & 3.1880 & 2.5173 & & h i & 9229.01 & p9 & 3d+ 2d & 9f+ 2f * & 18 & & 0.69 + 9239.58 & 0.0120 & 0.0095 & & ne ii & 9241.44 & & 2d & 2@xmath403@xmath41 * & 6 & 6 & 33.33 + 9259.11 & 0.0130 & 0.0103 & & o ii & 9260.50 & & 5d 4p & 4f p1 * & 2 & 2 & 26.92 + 9264.37 & 0.0230 & 0.0181 & & n ii & 9266.62 & v65.05a & 4d 3d * & 5f 1@xmath403@xmath41 & 5 & 7 & 17.39 + 9270.67 & 0.0090 & 0.0071 & & ne i & 9272.40 & & 3/2@xmath402/1@xmath41 * & 3/2@xmath403/2@xmath41 & 1 & 3 & 33.33 + 9274.33 & 0.0120 & 0.0095 & & fe i & 9276.23 & & z5i * & e5h & 11 & 9 & 29.17 + 9281.26 & 0.0130 & 0.0102 & & fe ii & 9283.39 & & c4f & z4 g * & 8 & 8 & 26.92 + 9287.81 & 0.0130 & 0.0102 & & [ fe iv ] & 9289.40 & & 4d & 2d5 & 8 & 6 & 15.38 + 9295.26 & 0.0100 & 0.0079 & & n ii & 9297.34 & v65.05a & 4d 3d * & 5f 1@xmath403@xmath41 & 7 & 7 & 30.00 + 9301.93 & 0.0080 & 0.0063 & & he i & 9303.19 & 3.9 & 3p 1p * & 9d 1d & 3 & 5 & 37.50 + 9304.75 & 0.0090 & 0.0071 & & o ii & 9306.56 & & 4d 4d & 5f d3 * & 6 & 8 & 33.33 + 9321.14 & 0.0150 & 0.0118 & & fe ii & 9323.06 & & e4d & 5d 5p4p * & 4 & 6 & 22.00 + 9329.70 & 0.0220 & 0.0173 & & ne ii & 9331.50 & & 5 g 2@xmath404@xmath41 & 8f 2@xmath405@xmath41 * & 8 & 10 & 15.00 + 9337.51 & 0.0130 & 0.0102 & & n ii & 9339.40 & & ? & ? & & & 25.38 + 9343.02 & 0.1080 & 0.0850 & & he ii & 9344.93 & 5.8 & 5g+ 2 g & 8h+ 2h * & 50 & & 4.63 + 9348.76 & 0.0170 & 0.0134 & & ne ii & 9350.15 & & 5 g 2@xmath406@xmath41 & 8f 2@xmath405@xmath41 * & 12 & 10 & 17.65 + 9364.92 & 0.0170 & 0.0134 & & he ii & 9367.03 & & 6f 2f * & 17 g 2 g & 32 & & 17.65 + 9384.41 & 0.0120 & 0.0094 & & ni ii & 9386.10 & & 4f 4@xmath402@xmath41 * & 3f 6d4p & 6 & 4 & 20.83 + 9390.01 & 0.0100 & 0.0079 & & [ fe iv ] & 9391.70 & & 3d5 2i & 3d5 2g2 & 12 & 10 & 25.00 + 9395.83 & 0.0160 & 0.0126 & & ni ii & 9397.77 & & 3f 5d4 g & 3f 7p4f * & 12 & 10 & 15.62 + 9401.19 & 0.0140 & 0.0110 & & ne ii & 9402.17 & & 4f 2@xmath404@xmath41 * & 5 g 1@xmath405@xmath41 & 10 & 12 & 14.29 + 9407.04 & 0.0130 & 0.0102 & & ne i & 9408.85 & & 3d 2@xmath404@xmath41 * & 8f 2@xmath404@xmath41 & 9 & 9 & 23.08 + & & & * & o
ii & 9409.06 & & 4d 4p & 5f d2 * & 4 & 6 & + 9415.59 & 0.0120 & 0.0094 & & ne ii & 9417.43 & & 6 g 2@xmath402@xmath41 & 4f2@xmath403@xmath41 * & 4 & 6 & 16.67 + 9423.47 & 0.0210 & 0.0165 & & ne i & 9425.38 & v8.01 & 3p 2@xmath401@xmath41 & 3d 2@xmath401@xmath41 * & 1 & 3 & 11.90 + 9434.65 & 0.0320 & 0.0251 & & [ fe ii ] & 9436.63 & & ( 3f2)4s 2f & ( 1d2)4s 2d & 6 & 4 & 12.50 + 9448.11 & 0.0140 & 0.0110 & & ne ii & 9449.81 & & 4d 4p & 5f 1@xmath403@xmath41 * & 6 & 8 & 25.00 + 9451.88 & 0.0070 & 0.0055 & & ne i & 9453.98 & & 3d 1@xmath402@xmath41 * & 8f1@xmath403,4@xmath41 & 5 & 7 & : + 9458.25 & 0.0170 & 0.0133 & & ne ii & 9459.07 & & 4d 2f & 5f 2@xmath405@xmath41 * & 8 & 10 & 17.65 + 9462.28 & 0.0940 & 0.0737 & & he i & 9463.58 & v67 & 3s 3s & 5p 3p * & 3 & 9 & 4.26 + 9472.60 & 0.0160 & 0.0125 & & ni ii & 9475.03 & & 4f 3@xmath403@xmath41 * & 3f 6d2 g & 8 & 10 & 21.88 + 9476.97 & 0.0350 & 0.0274 & & [ co iv ] & 9479.00 & & 3f2 & 1i & 9 & 13 & 11.43 + 9485.10 & 0.0200 & 0.0157 & & [ fe ii ] & 9486.98 & & b4p & b4d & 6 & 2 & 17.50 + 9489.22 & 0.0100 & 0.0078 & & o ii & 9490.85 & & 4d 4p & 5f d3 * & 6 & 8 & 35.00 + 9498.01 & 0.0090 & 0.0071 & & ne ii & 9499.27 & & 4d 2f & 5f 2@xmath403@xmath41 * & 8 & 8 & 44.44 + 9515.26 & 0.0790 & 0.0619 & & he i & 9516.63 & v76 & 3p 3p * & 7d 3d & 9 & 15 & 6.33 + 9529.21 & 55.7150 & 43.6169 & & [ s iii ] & 9530.60 & & 3p2 3p & 3p2 1d & 5 & 5 & 1.44 + 9543.97 & 1.8760 & 1.4680 & & h i & 9545.97 & p8 & 3d+ 2d & 8f+ 2f * & 18 & & 1.33 + 9560.71 & 0.0260 & 0.0203 & & ne i & 9562.65 & & 3/2@xmath401/2@xmath41 & 3/2@xmath403/2@xmath41 * & 3 & 3 & 19.23 + 9574.04 & 0.0170 & 0.0133 & & ar ii & 9576.17 & & 4f 2@xmath403@xmath41 * & 5 g 2@xmath402@xmath41 & 6 & 6 & 29.41 + 9577.48 & 0.0130 & 0.0102 & & [ mn iii ] & 9579.60 & & b4p & d2 g & 6 & 8 & 30.77 + 9584.52 & 0.0180 & 0.0141 & & ni ii & 9586.59 & & 4f 4@xmath405@xmath41 * & 3f 6d4h & 12 & 14 & 19.44 + 9596.54 & 0.0100 & 0.0078 & & c iii & 9597.81 & v17.01 & 4f 1f * & 3p 1d & 7 & 5 & 30.00 + & & & * & fe ii & 9598.78 & & e4d & 5d 5p6f * & 6 & 8 & + 9601.65 & 0.0260 & 0.0203 & & he i & 9603.44 & v71 & 3s 1s & 6p 1p * & 1 & 3 & 11.54 + 9607.80 & 0.0100 & 0.0078 & & fe i & 9609.45 & & e7d & 4f 5p3d * & 7 & 7 & 30.00 + 9612.63 & 0.0140 & 0.0109 & & ne ii & 9614.55 & & 4d 2f & 5f 1@xmath403@xmath41 * & 6 & 8 & 17.86 + 9618.77 & 0.0090 & 0.0070 & & c iii & 9620.50 & & 3h * & 3 g & 33 & 27 & 27.78 + 9624.25 & 0.0210 & 0.0164 & & he i & 9625.70 & v90 & 3p 1p * & 8d 1d & 3 & 5 & 14.29 + 9630.37 & 0.0150 & 0.0117 & & ne ii & 9632.33 & & 4d 2d & 5f 2@xmath402@xmath41 * & 4 & 6 & 16.67 + 9632.46 & 0.0130 & 0.0101 & & fe ii & 9634.21 & & e6d & 4p sp6d * & 8 & 10 & 15.38 + 9643.00 & 0.0150 & 0.0117 & & [ fe iii ] & 9645.10 & & 5p & 3d & 3 & 7 & 16.67 + 9651.86 & 0.0330 & 0.0257 & & [ fe ii ] & 9652.70 & & a4f & a4p & 4 & 6 & 7.58 + 9658.45 & 0.0100 & 0.0078 & & ar i & 9660.70 & & 3/2@xmath401/2@xmath41 & 3/2@xmath403/2@xmath41 * & 3 & 3 & 23.00 + 9668.10 & 0.0130 & 0.0101 & & [ fe ii ] & 9669.66 & & ( 3p2)4s 2p & ( 3d)4s 2d & 4 & 4 & 17.69 + 9678.65 & 0.0110 & 0.0086 &
& c i & 9680.62 & & 3d 3d * & 7f 2@xmath402@xmath41 & 7 & 5 & 22.73 + 9690.51 & 0.0160 & 0.0125 & & o i & 9693.00 & & ? & ? & & & 18.75 + 9699.74 & 0.0350 & 0.0273 & & he i & 9702.71 & v75 & 3p 3p * & 7s 3s & 9 & 3 & 10.00 + 9706.31 & 0.0130 & 0.0101 & & c iii & 9708.20 & & 6h 3h * & 7i 3i & 33 & 39 & 23.08 + 9711.61 & 0.0170 & 0.0132 & & ar i & 9713.99 & & 3d 2@xmath402@xmath41 * & 8p 2@xmath401@xmath41 & 3 & 1 & 17.65 + 9718.03 & 0.0140 & 0.0109 & & ne ii & 9720.13 & & 4p 2s * & 4d 2p & 2 & 2 & 17.86 + 9739.64 & 0.0100 & 0.0078 & & o i & 9741.50 & v55.04 & 3p 1f & 3d 1 g * & 7 & 9 & 24.00 + 9746.19 & 0.0130 & 0.0101 & & o i ] & 9748.40 & & 5p & 3d * & 5 & 15 & 18.46 + 9760.45 & 0.0530 & 0.0412 & & he ii & 9762.15 & & 6f 2f * & 15 g 2 g & 32 & & 7.55 + 9788.41 & 0.0120 & 0.0093 & & fe ii & 9790.52 & & 4f 4@xmath406@xmath41 * & 5 g 4@xmath407@xmath41 & 14 & 16 & 33.33 + 9791.64 & 0.0140 & 0.0109 & & n ii & 9794.01 & v65.14 & 4d 1f * & 5f 2@xmath405@xmath41 & 7 & 9 & 25.00 + 9797.36 & 0.0090 & 0.0070 & & ne ii & 9799.26 & & 2p * & 2p & 4 & 4 & 38.89 + 9806.89 & 0.0070 & 0.0054 & & ne ii & 9808.85 & & 4s 2p & 4p 2p * & 2 & 4 & 50.00 + 9814.67 & 0.0060 & 0.0047 & & fe ii & 9816.40 & & y2h * & 4s2 2 g & 10 & 8 & 58.33 + 9822.25 & 0.0100 & 0.0078 & & [ c i ] & 9824.13 & & 2p2 3p & 2p2 1d & 3 & 5 & 33.00 + 9831.74 & 0.0050 & 0.0039 & & o ii & 9833.39 & & d@xmath402@xmath41 * & 2@xmath402@xmath41 & 6 & 10 & : + 9848.24 & 0.0360 & 0.0279 & & [ c i ] & 9850.26 & & 2p2 3p & 2p2 1d & 5 & 5 & 9.17 + & & & * & o ii & 9849.96 & & 4d 2d & 5f f3 * & 4 & 6 & + 9866.10 & 0.0120 & 0.0093 & & n ii & 9868.21 & v65.09 & 4f 1@xmath403@xmath41 & 5 g 1@xmath404@xmath41 * & 5 & 7 & 29.17 + 9888.75 & 0.0120 & 0.0093 & & n ii & 9891.09 & v65.11b & 4f 1@xmath404@xmath41 & 5 g 1@xmath405@xmath41 * & 18 & 27 & 26.67 + 9901.82 & 0.2650 & 0.2054 & & c ii & 9903.46 & v17.02 & 4f 2f * & 5 g 2 g & 22 & 26 & 3.64 + & & & * & o ii & 9903.39 & & 4d 2d & 5f f3 * & 6 & 6 & + & & & * & o ii & 9903.45 & & 4d 2d & 5f f3 * & 6 & 8 & + 9909.63 & 0.0060 & 0.0047 & & o ii & 9911.46 & & 2@xmath404@xmath41 & h@xmath405@xmath41 * & 10 & 10 & : + 9961.60 & 0.0380 & 0.0294 & & fe ii & 9963.77 & & 4f 1@xmath403@xmath41 * & 5 g 1@xmath404@xmath41 & 6 & 8 & 13.16 + 9966.65 & 0.0150 & 0.0116 & & n ii & 9969.34 & v65.16a & 4f 2@xmath404@xmath41 & 5 g 2@xmath405@xmath41 * & 9 & 11 & 21.33 + & & & * & n ii & 9969.43 & v65.16a & 4f 2@xmath404@xmath41 & 5 g 2@xmath405@xmath41 * & 9 & 9 & + 9980.62 & 0.0320 & 0.0248 & & o ii & 9982.42 & & 4f g5 * & 5 g 2(6 ) & 12 & 14 & 13.44 + 9988.51 & 0.1520 & 0.1176 & & o ii & 9990.27 & & 4f d2 * & 5 g 1(3 ) & 6 & 8 & 6.58 + & & & * & o ii & 9990.27 & & 4f d3 * & 5 g 0(4 ) & 6 & 8 & + 9991.12 & 0.0330 & 0.0255 & & c ii & 9993.30 & & 4s & 4p * & 4 & 4 & 30.30 + 10002.64 & 0.0230 & 0.0255 & & o ii & 10004.46 & & g@xmath404@xmath41 * & 1@xmath403@xmath41 & 10 & 8 & 43.48 + 10008.47 & 0.0570 & 0.0255 & & ne ii & 10010.22 & & 4d 4p & 5f 2@xmath402@xmath41 * & 4 & 4 & 17.54 + 10012.53 & 0.2180 & 0.0255 & & [ fe ii ] & 10013.89 & & 3d7 2 g & ( 3h)4s 2h & 8 & 10 & 5.87 + 10020.79 & 0.0800 & 0.0255 & & he i & 10023.10 & & 3d 1d & 7p 1p * & 5 & 3 & 12.50 + 10026.10 & 0.3120 & 0.2411 & & he i & 10027.72 & v81 & 3d 3d & 7f 3f * & 15 & 21 & 3.53 + 10028.19 & 0.0130 & 0.2411 & & ne ii & 10030.20 & & 4f 2@xmath403@xmath41 * & 5 g 2@xmath405@xmath41 & 8 & 10 & 38.46 + 10037.08 & 0.1310 & 0.2411 & & o ii & 10038.70 & & 3d7 2h & ( 3g)4s 2 g & 10 & 8 & 8.40 + 10047.60 & 5.7340 & 4.4294 & & h i & 10049.37 & p7 & 3d+ 2d & 7f+ 2f * & 18 & & 1.08 + 10081.76 & 0.1530 & 4.4294 & & ne ii & 10083.80 & & 4d 4p & 5f 2@xmath402@xmath41 * & 6 & 6 & 9.80 + 10096.17 & 0.0830 & 4.4294 & & [ fe ii ] & 10098.09 & & b2h & b2d & 10 & 6 & 15.66 + 10108.75 & 0.0970 & 4.4294 & & ne ii & 10110.70 & & 4p 2p * & 5s 2p & 4 & 4 & 13.40 + 10121.80 & 3.7810 & 2.9155 & & he ii & 10123.61 & 4.5 & 4f+ 2f * & 5g+ 2 g & 32 & 50 & 1.80 + 10158.39 & 0.0610 & 2.9155 & & cr ii & 10160.60 & & 5d 4f4d * & 5 g 0@xmath404@xmath41 & 6 & 8 & 18.85 + 10189.87 & 0.0750 & 2.9155 & & ne ii & 10191.73 & & 2d & 2d * & 6 & 4 & 16.00 + 10201.94 & 0.0880 & 2.9155 & & ca i & 10203.60 & & 4s4d 3d & 4s8f 1f * & 7 & 7 & 16.48 + 10210.39 & 0.1070 & 2.9155 & & [ cr ii ] & 10212.11 & & a4 g & b4p & 8 & 4 & 18.69 + 10290.43 & 0.0990 & 2.9155 & & o ii & 10292.08 & & g@xmath405@xmath41 * & 0@xmath404@xmath41 & 10 & 18 & 17.68 + 10309.19 & 0.0990 & 2.9155 & & he i & 10311.00 & v74 & 3p 3p * & 6d 3d & 9 & 15 & 17.68 + 10372.08 & 0.2240 & 2.9155 & & n ii & 10374.40 & & ? & ? & & & 11.38 + 10387.62 & 0.0700 & 2.9155 & & n ii & 10389.80 & & 3p * & 5p & 5 & 7 & 36.43 + 10402.21 & 0.0950 & 2.9155 & & si i & 10404.00 & & 3d 3d * & 7f 2@xmath402@xmath41 & 3 & 5 & 42.63 + 10418.60 & 0.4490 & 0.3439 & & ne ii & 10420.71 & & 3d 4p & 4p 4p * & 4 & 4 & 17.82 + 10469.32 & 0.2290 & 0.3439 & & fe i & 10471.30 & & w5f * & f5p & 3 & 3 & 26.20 + 10489.56 & 0.0710 & 0.3439 & & o ii & 10491.40 & & 4f d3 * & 5d 4d & 8 & 6 & : + 10528.42 & 0.1960 & 0.3439 & & o ii & 10530.70 & & 4f f2 * & 5d 2f & 6 & 8 & 42.86 + 10620.70 & 0.1650 & 0.1258 & & n i & 10623.00 & v52 & 3p 4p * & 3d 4p & 2 & 2 & 51.52 +
x. fang and x .- w .
liu thank p. j. storey for making the o ii effective recombination coefficients available prior to publication .
99 aggarwal , k. m. , 1983 , apjs , 52 , 387 aller , l. h. , 1941 , apj , 93 , 236 aller , l. h. , 1961 , the abundances of the elements ( new york , interscience ) , p. 71 aller , l. h. , kaler , j. b. , 1964a , apj , 139 , 1074 aller , l. h. , kaler , j. b. , 1964b , apj , 140 , 621 aller , l. h. , menzel , d. h. , 1945 , apj , 102 , 239 baldwin , j. a. , verner , e. m. , verner , d. a. , ferland , g. j. , martin .
p. g. , korista , k. t. , rubin , r. h. , 2000 , apjs , 129 , 229 balick , b. , alexander , j. , hajian , a. r. , terzian , y. , perinotto , m. , patriarchi , p. , 1998 , aj , 116 , 360 baluja , k. l. , zeippen , c. j. , 1988 , j. phys .
b : at . mol .
, 21 , 1455 barker , t. , 1983 , apj , 267 , 630 benjamin , r. a. , skillman , e. d. , smits , d. p. , 1999 , apj , 514 , 307 berman , l. , 1930 , lick obs . bull . , 15 , no .
430 , 86 berrington , k. a. , 1988 , j. phys .
phys . , 21 , 1083 bowen , i. s. , wyse , a. b. , 1939 , lick obs .
, 19 , no . 495 , 1 brocklehurst , m. 1972 , mnras , 157 , 211 butler , k. , zeippen , c. j. , 1994 , a&as , 108 , 1 esteban , c. , peimbert , m. , torres - peimbert , s. , escalante , v. , 1998 , mnras , 295 , 401 esteban , c. , peimbert , m. , torres - peimbert , s. , garcia - rojas , f. , rodriguez , m. , 1999 , apj , 120 , 113 fang , x. , liu , x .- w . , 2010 , in preparation hirata , r. , horaguchi , t. , 1995 , atomic spectral line list , ( vizier on - line data catalogue ) howarth , i. d. , 1983 , mnras , 203 , 301 johnson , c. t. , kingston , a. e. , 1990 , j. phys .
b , 23 , 3393 kaler , j. b. , aller , l. h. , 1969 , apj , 157 , 1231 keenan , f. p. , aller , l. h. , bell , k. l. , hyung , s. , mckenna , f. c. , ramsbottom , c. a. , 1996 , mnras , 281 , 1073 keenan , f. p. , hibbert , a. , ojha , p. c. , conlon , e. s. , 1993 , phys .
, 48 , 129 kingsburgh , r. l. , barlow , m. j. , 1992 , mnras , 257 , 317 liu , y. , liu , x .- w . , luo , s .-
barlow , m. j. , 2004a , mnras , 353 , 1231 liu , y. , liu , x .- w . ,
barlow , m. j. , luo , s .-
, 2004b , mnras , 353 , 1251 liu , x .- w . , danziger , i. j. , 1993 , mnras , 263 , 256 liu , x .- w . , storey , p. j. , barlow , m. j. , clegg , r. e. s. , 1995 , mnras , 272 , 369 liu , x .- w . , storey , p. j. , barlow , m. j. , danziger , i. j. , cohen , m. , bryce , m. , 2000 , mnras , 312 , 585 liu , x .- w . , luo , s .-
barlow , m. j. , danziger , i.j . ,
storey , p. j. , 2001 , mnras , 327 , 141 liu x .- w . , 2003 , in kwok s. , dopita m. , sutherland r. , eds , iau symp .
@xmath43209 , planetary nebulae : their evolution and role in the universe .
, san francisco , p. 339
liu , x .- w . , barlow , m. j. , zhang , y. , bastin , r j. , storey , p. j. , 2006a , mnras , 368 , 1959 liu , x .- w .
, 2006 , in barlow , m. j. , mndez , r. h. , eds , iau symp , @xmath43234 , planetary nebulae in our galaxy and beyond , proceedings of the iau symposium .
cambridge : cambridge university press , p. 219
liu , x .- w . ,
barlow , m. j. , 2001 , mnras,326 , 1049 mendez , r. h. , kudritzki , r. p. , herrero , a. , 1992 , a&a , 260 , 329 mendoza , c. 1983 , in iau symp .
103 , planetary nebulae , ed .
flower , d. , reidel , dordrecht , p.143 mendoza , c. , zeippen , c. j. , 1982a , mnras , 199 , 1025 mendoza , c. , zeippen , c. j. , 1982b , mnras , 198 , 127 mendoza , c. , zeippen , c. j. , 1983 , mnras , 202 , 981 nahar , s. n. , pradhan , a. k. , 1996 , a&as , 119 , 509 nussbaumer , h. , rusca , c. , 1979 , a&a , 72 , 129 nussbaumer , h. , storey , p. j. , 1981 , a&a , 99 , 177 peimbert , m. , 1967 , apj , 150 , 825 pquignot , d. , baluteau , j. p. , 1994 ,
a&a , 283 , 593 perinotto , m. , benvenuti , p. , 1981
, a&a , 101 , 88 pradhan , a. k. , 1976 , mnras , 177 , 31 rugin , r. h. , 1986 , apj , 309 , 334 rubin , r. h. , 1989 , apjs , 69 , 897 sawey , p. m. j. , berrington , k. a. , 1993 , at .
data nucl .
data tables , 55 , 81 sharpee , b. , baldwin , j. a. , williams , r. , 2004 , apj , 615 , 323 sharpee , b. , williams , r. , baldwin , j. a. , van hoof , p. a. m. , 2003 , apjs , 149 , 157 sharpee , b. , zhang , y. , williams , r. , pellegrini , e. , cavagnolo , k. , baldwin , j. a. , phillips , m. , liu , x .- w .
, 2007 , apj , 659 , 1265 smits , d. p. , 1996 ,
mnras , 278 , 683 stafford , r. p. , bell , k. l. , hibbert , a. , wijesundera , w. p. , 1994 ,
mnras , 268 , 816 storey , p. j. , hummer , d. g. , 1995 , mnras , 272 , 41 tsamis , y. g. , barlow , m. j. , liu , x .- w . , danziger , i. j , storey p. j. , 2003 , mnras , 345 , 186 tsamis , y. g. , barlow , m. j. , liu , x .- w . , storey , p. j. , danziger , i. j. , 2004 , mnras , 353 , 953 viegas , s. , clegg , r. e. s. , 1994 , mnras , 271 , 993 wang , w. , liu , x .- w . , 2007 , mnras , 381 , 669 wyse , a. b. , 1942 ,
apj , 95 , 356 zhang , y. , liu , x .- w . , luo , s .-
g . , pquignot , d. , barlow , m. j. , 2005a , a&a , 442 , 249 zhang , h. l. , 1996 , a&as , 119 ,
523 zhang , y. , liu , x .- w . , liu , y. , rubin , r. h. , 2005b , mnras , 358 , 457 zhang , y. , liu , x .- w . ,
wesson , r. , storey , p. j. , liu , y. , danziger , i. j. , 2004 , mnras , 351 , 935 zhang , y. , yuan , h .- b .
, hua , c .- t . ,
liu , x .- w . , nakashima , j. , kwok , s. , 2009 , apj , 695 , 488 zeippen , c. j. , 1982 , mnras , 198 , 111 zeippen , c. j. , le bourlot , j. , butler , k. , 1987 , a&a , 188 , 251 | we present very deep ccd spectrum of the bright , medium - excitation planetary nebula ngc7009 , with a wavelength coverage from 3040 to 11,000 .
traditional emission line identification is carried out to identify all the emission features in the spectra , based on the available laboratory atomic transition data .
since the spectra are of medium resolution , we use multi - gaussian line profile fitting to deblend faint blended lines , most of which are optical recombination lines ( orls ) emitted by singly ionized ions of abundant second - row elements such as c , n , o and ne .
computer - aided emission - line identification , using the code emili developed by sharpee et al .
, is then employed to further identify all the emission lines thus obtained . in total about 1200 emission features
are identified , with the faintest ones down to fluxes 10@xmath0 of h@xmath1 .
the flux errors for all emission lines , estimated from multi - gaussian fitting , are presented .
plots of the whole optical spectrum , identified emission lines labeled , are presented along with the results of multi - gaussian fits of all the properly identified emission lines , permitted lines contribute 81 per cent to the total line number .
more than 200 o ii permitted lines are presented , as well as many others from n ii and ne ii . due to its relatively simple atomic structure ,
c ii presents few lines . within the flux range 10@xmath2
10@xmath0 h@xmath1 where most permitted lines of c ii , n ii , o ii and ne ii fall , the average flux measurement uncertainties are about 10 to 20 per cent .
comparison is also made of the number of emission lines identified in the current work of ngc7009 and those of several other planetary nebulae ( pne ) that have been extensively studied in the recent literature , and it shows that our line - deblending procedure increases the total line number significantly , especially for emission lines with fluxes lower than 10@xmath3 of h@xmath1 .
higher resolution is still needed to obtain more reliable fluxes for those extremely faint emission lines , lines of fluxes of the order of 10@xmath4 10@xmath5 of h@xmath1 .
plasma diagnostics using optical forbidden line ratios give an average electron temperature of 10,020 k , which agrees well with previous results of the same object .
the average electron density of ngc7009 derived from optical forbidden line ratios is 4290 cm@xmath3 .
the [ o iii ] @xmath74959/@xmath74363 nebular - to - auroral line ratio yields an electron temperature of 9800 k. the ratio of the nebular continuum balmer discontinuity at 3646 to h 11 reveals an electron temperature of 6500 k , about 600 k lower than the measurements published in the literature .
the balmer decrement reveals a density of about 3000 cm@xmath3 .
also derived are electron temperatures from the he i line ratios , and a value of 5100 k from the @xmath77281/@xmath76678 ratio is adopted . utilizing the effective recombination coefficients newly available
, we find an electron temperature around 1000 k from o ii orl spectrum .
thus general pattern of electron temperatures , @xmath8([o iii ] ) @xmath9 @xmath8(h i bj ) @xmath9 @xmath8(he i ) @xmath9 @xmath8(o ii ) , which is seen in many pne , is repeated in ngc7009 .
far - ir fine - structure lines , with observed fluxes adopted from the literature , are also used to derive @xmath8 and @xmath10 .
the [ o iii ] ( 52@xmath11 m + 88@xmath11m)/@xmath74959 line ratio gives an electron temperature of 9260 k , and the 52@xmath11m/88@xmath11 m ratio yields an electron density of 1260 cm@xmath3 .
[ firstpage ] line : identification atomic data atomic processes planetary nebulae : individual : ngc7009 |
ultraluminous x - ray sources ( ulxs ) are variable off - nuclear x - ray sources with luminosities exceeding the eddington luminosity of a @xmath4 compact object , assuming isotropic emission @xcite . irregular variability , on time scales from seconds to years , suggests that ulxs contain accreting compact objects .
intermediate mass black holes would be required to produce the inferred luminosities , but ulxs may , instead , accrete at super eddington rates or be beamed , mechanically or relativistically .
ngc 5408 x-1 is one of the best intermediate mass black hole candidates because it powers a radio nebula requiring an extremely energetic outflow @xcite and a photoionized nebula requiring an x - ray luminosity above @xmath5 @xcite .
also , quasi - periodic x - ray oscillations at low frequencies suggest a high compact object mass @xcite .
the optical counterpart to ngc 5408 x-1 was identified by @xcite and optical spectra were obtained by @xcite .
the optical spectra had no absorption lines suggesting the emission is not dominated by the companion star .
the observed continuum emission may arise from a nebula or reprocessing of x - rays in an accretion disk .
the optical spectrum is dominated by emission lines , including forbidden lines which must be produced in a low density environment such as a nebula .
several high excitation lines were detected indicating that the nebula is x - ray photoionized .
@xcite found that the heii line from ngc 5408 x-1 was broader than the forbidden lines .
permitted lines produced in the high - density environment of an accretion disk can be broad , reflecting the distribution of velocities within the optical emitting regions of the disk .
furthermore , since the accretion disk moves with the compact object , the line velocity shifts may provide a means to constrain the compact object mass @xcite . to study the heii line profile of ngc 5408 x-1 in more detail
, we obtained new observations using the fors-2 spectrograph on the european southern observatory very large telescope ( vlt ) with a high resolution grism and reanalyzed our previous fors-1 observations @xcite . the observations and data reduction
are described in 2 .
the results are presented in 3 and discussed in 4 .
fors-2 observations of ngc 5408 x-1 were obtained on 12 april 2010 using the gris_1200b and gris_1200r grisms with a slit width of @xmath6 covering the spectral range 3660@xmath75110 and 5750@xmath77310 with dispersion 0.36 pixel@xmath8 and 0.38 pixel@xmath8 and spectral resolution @xmath9 and @xmath10 at the central wavelength , respectively . the observation block ( ob ) consisted of three 849 s exposures with a 12 pixel offset along the spatial axis between successive exposures .
ccd pixels were binned for readout by 2 in both the spatial and spectral dimensions .
we also reanalyzed all six obs of the previous fors-1 observations @xcite , hereafter the low resolution data ( lrd ) , taken using the gris_600b grism which has a spectral resolution of @xmath11 at the central wavelength and with three shifted exposures per ob .
the average seeing for our new observations was 0.72 and 0.62 arcsecond for the blue and red spectra , respectively .
the average seeing of the six obs of the lrd were 0.87 , 0.82 , 0.96 , 1.28 , 0.64 , 0.57 arcsecond , respectively .
data reduction was carried out using the image reduction and analysis facility ( iraf ) @xcite .
first , we created bias and flat - field images , then applied these to correct the spectrum images .
the three exposures in each ob were aligned then averaged to eliminate bad pixels and cosmic rays using the imcombine task with the ccdclip rejection algorithm .
as the continuum emission of the ulx counterpart is faint , we could not trace its spectrum .
following @xcite , we used the bright nearby star at 2mass @xcite position @xmath12 as a reference trace .
the trace position on the spatial axis varied less than half a pixel along the whole length of the dispersion axis .
the trace for the ulx counterpart was centered on the heii @xmath13 emission line profile .
the smallest possible trace width , 2 pixels corresponding to @xmath14 , was used to best isolate the ulx emission from the nebular emission .
background subtraction was done with a trace close by . the hgcdhenea lamp and standard star ltt7379
were used for wavelength and flux calibration .
an atmospheric extinction correction was applied using the iraf built - in cerro tololo inter - american observatory ( ctio ) extinction tables . to estimate the reddening
, we used the balmer decrement of h@xmath15/h@xmath1 , we find @xmath16 in agreement with @xcite . we corrected for reddening using the extinction curve from @xcite with @xmath17 . to study the kinematics , we need to characterize the instrumental resolution in order to obtain intrinsic line widths . after applying the dispersion correction to the lamp spectrum
, we measured the full width at half maximum ( fwhm ) of several lines , excluding saturated ones , by fitting gaussians with the iraf splot subroutine .
the instrumental fwhm was 2.24 and 5.08 for the high and low resolution data , respectively .
the error on the instrumental fhwm was estimated by finding the standard deviation of the fwhm for several different lines .
for the heii , h@xmath1 and [ oiii ] emission lines , see fig . 1 . and table [ fit ] , we first fitted the continuum with a second order polynomial to a region around each line excluding the line itself by visual examination .
we estimated the measurement errors by calculating the root mean square deviation of the data in the same region .
then , we performed a non - linear least squares fit using the lmfit subroutine of the interactive data language version 7.0 and based on `` mrqmin '' @xcite .
we fitted the line profiles iteratively , first using one gaussian which converged on the narrow component , then using a sum of two gaussians with initial parameters adjusted to achieve convergence .
all six parameters in the two gaussian fit were free to vary .
the errors on the parameters were calculated by the fitting routine in a way that the uncertainty for the @xmath18th parameter derives from the square - root of the corresponding diagonal element of the covariance matrix .
the intrinsic line width was calculated assuming the measured line width is the quadrature sum of the intrinsic and instrumental widths and the error on the intrinsic line width included a term for the uncertainty in the instrumental fhwm . for the h@xmath19 line , we fitted the sum of four gaussians , because the [ nii ] lines lie on the red and blue parts of the line wing .
initial fits to the h@xmath19 and red [ nii ] lines provided initial values for a fit with four gaussians . because the blue [ nii ] line has very low signal to noise ratio , the widths of the two [ nii ] lines were set equal , the wavelength offset was fixed at @xmath20 , and the amplitude of the blue line
was set to 1/3 of the red line @xcite .
[ new ] and [ o iii ] lines from the new , high resolution data
. results of line fits with two gaussian components are shown as the separate components plotted with dashed - dot lines and the overall fit plotted with dashed lines ( except [ o iii]).,title="fig:",width=336 ] and [ o iii ] lines from the new , high resolution data .
results of line fits with two gaussian components are shown as the separate components plotted with dashed - dot lines and the overall fit plotted with dashed lines ( except [ o iii]).,title="fig:",width=336 ] and [ o iii ] lines from the new , high resolution data .
results of line fits with two gaussian components are shown as the separate components plotted with dashed - dot lines and the overall fit plotted with dashed lines ( except [ o iii]).,title="fig:",width=321 ] [ lrd ]
the improved resolution of the new data clearly resolves a broad component in the heii line profile , see fig . 1 . and table [ fit ] .
the centroid is shifted from the nebular component by @xmath21 in the red direction .
we also searched for broad components in other lines .
h@xmath1 has a broad component with a fwhm similar to the heii line but shifted by @xmath22 towards the blue , rather than the red .
in contrast , a single gaussian provides a good fit to the forbidden [ oiii ] line and there is no evidence for a broad component , as expected if the line is emitted only from the nebula .
then we fitted the heii line profiles of the six obs of the lrd , see fig . 2 . and
table [ fit ] .
the flux variation of the overall line profiles correlates with the seeing , e.g. ob5 has the best seeing and the highest flux .
we detected a broad component in the heii line in ob3 , ob5 , and ob6 .
we did not significantly detect a broad component in ob1 , ob2 and ob4 .
this may be due to seeing or variations in the flux of the broad component .
lccccc line & wavelength ( ) & fwhm ( ) & flux & velocity ( km s@xmath8 ) & @xmath23 + heii & 4693.75 @xmath24 0.04 & 2.54 @xmath24 0.14 & 3.39 @xmath24 0.13 & 76.3 @xmath24 3.4 & 0.94 + & 4694.62 @xmath24 0.26 & 11.98 @xmath24 0.74 & 1.19 @xmath24 0.10 & 752.0 @xmath24 46.4 & + h@xmath1 & 4869.73 @xmath24 0.01 & 2.28 @xmath24 0.05 & 8.37 @xmath24 0.11 & 25.2 @xmath24 3.3 & 1.08 + & 4869.21 @xmath24 0.32 & 12.49 @xmath24 0.94 & 0.71 @xmath24 0.07 & 756.9 @xmath24 57.5 & + @xmath25o ] @xmath264959 & 4967.52 @xmath24 0.01 & 2.48 @xmath24 0.03 & 8.42 @xmath24 0.08 & 64.6 @xmath24 1.7 & 1.08 + heii ob3 & 4693.90 @xmath24 0.15 & 5.15 @xmath24 0.52 & 1.34 @xmath24 0.19 & 52.0 @xmath24 33.3 & 1.04 + & 4692.08 @xmath24 0.77 & 11.09 @xmath24 1.44 & 0.56 @xmath24 0.20 & 630.5 @xmath24 91.8 & + heii ob5 & 4693.80 @xmath24 0.12 & 5.81 @xmath24 0.41 & 1.74 @xmath24 0.17 & 179.8 @xmath24 25.7 & 1.03 + & 4690.03
@xmath24 1.43 & 11.75 @xmath24 1.33 & 0.49 @xmath24 0.13 & 678.0 @xmath24 84.7 & + heii ob6 & 4693.97 @xmath24 0.12 & 5.56 @xmath24 0.46 & 1.70 @xmath24 0.23 & 143.7 @xmath24 29.5 & 0.95 + & 4689.94 @xmath24 1.70 & 10.82 @xmath24 1.55 & 0.53 @xmath24 0.15 & 610.7 @xmath24 98.9 & + heii avg & 4693.89 @xmath24 0.08 & 5.50 @xmath24 0.26 & 1.58 @xmath24 0.10 & 135.0 @xmath24 16.6 & 1.15 + & 4690.82 @xmath24 0.68 & 11.47 @xmath24 0.73 & 0.53 @xmath24 0.09 & 657.5 @xmath24 46.3 & + we note that @xcite reported lower fluxes for heii , [ nev ] , and the continuum emission for ob4 ( with by far the worst seeing ) as compared to the other obs , while the other line fluxes remained relatively constant .
our new analysis suggests that this is due to changes in the seeing .
if the emitting region is smaller than the @xmath27 slit used for the lrd , then poor seeing will decrease the flux through the spectrometer .
if these emission components are enhanced close to the ulx system while the other line emission is uniform , then the poor seeing in ob4 would produce the observed changes in flux .
thus , there is no evidence for temporal variability of the continuum or line emission .
however , the subtraction performed by @xcite to isolate continuum emission arising from near the ulx is still justified ; the separation of components is spatial instead of temporal .
the heii line parameters are consistent between ob3 , ob5 , and ob6 .
the wavelength shifts of the heii broad component of ob3 , ob5 and ob6 relative to the narrow component are @xmath28 , @xmath29 @xmath30 and @xmath31 into the blue direction instead of the red as in the hrd , and are consistent within one @xmath32 .
we averaged the spectra for these three observations and fit the resulting line profile .
the fit results are listed as heii avg in table [ fit ] .
the shift of average line profile is @xmath33 .
the heii broad component width is consistent between the new and old data .
the narrow component is wider in the old data because we do not resolve the nebular lines .
the line fluxes are higher in the new data , most likely due to the wider slit .
the centroids of the narrow component are consistent , while the wavelength shift of the heii broad component between the old and new data is @xmath34 .
fitting the h@xmath1 line of the lrd , we did not get a good fit , due to the lack of spectral resolution and the low broad to narrow flux ratio .
we could not fit the bluer balmer lines because of their low s / n ratios .
we did fit the h@xmath19 line in the new data . although we do not obtain a good fit ( @xmath35 ) because of the complicated line profile ( i.e. the two [ nii ] lines lie on the red and blue wing of the h@xmath19 line ) , we find that there is a broad component with a width of 19 , while the width of the nebular component is 2.7 .
the [ nii ] lines are narrow , with a typical width of 3 , quantitatively supporting that the forbidden , nebular lines do not have broad components .
our new , high - resolution spectra show narrow nebular lines and broad components in the heii , h@xmath1 , and h@xmath19 lines . our previous , moderate - resolution spectra show a broad component in the heii line .
there is no broad component in the [ o iii ] nebular lines in either the new or old spectra .
there is still no sign of any absorption lines in the new spectra .
the broad components of both heii and h@xmath1 have widths @xmath0750 km / s , consistent with production in the accretion disk , and are roughly gaussian , instead of having p - cygni profiles that would indicate origin in a wind .
following @xcite , we estimate the size of the line - emitting region , @xmath36 , by assuming the line - emitting gas is in keplerian orbits around a compact object , thus @xmath37 . we find @xmath38 au , which for a mass of 10 m@xmath3 would give an upper limit of 3.4 r@xmath3 .
this is consistent with origin of the broad heii line in the accretion disk .
the broad line components are shifted relative to the narrow components . in the new data ,
the shifts are small compared to the line width , @xmath39 km / s for heii and @xmath40 km / s for h@xmath1 .
these shifts are consistent only at the @xmath41 level , which might indicate a difference in the spatial origin of the lines .
however , this is still consistent with production of both lines within the disk since random motions within the disk and variation between the emission regions could produce shifts that are small compared to the line widths , as observed . the central wavelength of the heii broad component shifts markedly between the odld and new data , @xmath34 @xmath30 or @xmath42 km / s .
this shift is a substantial fraction of the line width .
the shift could be due to random motion within the disk , differing viewing geometries @xcite , or orbital motion of disk ( and the compact object ) .
if the shifts in the broad component of the heii line are due to orbital motion , then this would provide a means to determine the orbital period and would also provide a measurement of the mass function for the secondary star .
thus , a program of monitoring ngc 5408 x-1 with high - resolution optical spectroscopic observations will be important in extending our understanding of the physical nature of this system . in this section
, we make some speculations based on interpretation of the shift in the broad component of the heii line as due to orbital motion .
one can express the mass function and the compact object mass , @xmath43 , in terms of the orbital period , @xmath44 , the velocity excursion , @xmath45 , and the companion mass , @xmath46 , as [ binary ] donor and the lower line is for a 20 m@xmath3 donor ( left vertical scale ) .
the orbital separation is shown as dotted lines with values read on the right vertical scale ; the upper line for an 120 m@xmath3 and the lower is for 20 m@xmath3 donor ( right vertical scale ) .
the two dashed lines show the upper limit on the size of the line - emitting region ; the upper line is for an 120 m@xmath3 donor and the lower line is for a 20 m@xmath3 donor ( right vertical scale ) .
the intercept of the dotted and dashed lines provides a limit on the period and consequently on the mass of the black hole.,title="fig:",width=336 ] where @xmath18 is the inclination angle and @xmath49 is the gravitational constant . from the shift of the heii line quoted above , we constrain the semi - amplitude of the radial velocity @xmath50 km / s .
thus , if the maximum mass of the companion and the orbital period are known , then eq . 2 leads to an upper bound on the mass of the compact object .
the binary system has a visual magnitude @xmath51 that gives an upper limit on the absolute magnitude of the companion of @xmath52 at a distance of 4.8 mpc @xcite .
unfortunately , this places little restriction on the companion mass as even o3v stars , with masses of 120 @xmath53 , are allowed . however , very high mass stars are very short lived , no more than a few million years .
there is no evidence of a dense stellar association near ngc 5408 x-1 and origin in the closest super - star cluster would require a transit time to the present location on the order of 30 myr @xcite .
thus , the companion mass is likely significantly lower , near 20 @xmath53 or less similar to found from studies of the stellar environments of other ulxs @xcite .
figure 3 . shows the upper bound on the compact object mass for donors of 120 m@xmath3 and 20 m@xmath3 as a function of orbital period .
high black hole masses are excluded , except for very short periods .
we note that the heii line shift was the same in ob3 versus ob5 and ob6 , taken one day apart , suggesting that the period is longer than few days .
thus , the black hole mass is likely below @xmath01800 m@xmath3 .
the more probable companion mass of 20 m@xmath3 or less would imply smaller black hole masses , less than 112 m@xmath3 . as a further constraint , we note that the orbital separation should be larger than the size of the emitting region , calculated above . assuming a circular orbit , the orbital separation of the compact object is @xmath54 .
figure 3 .
shows the the orbital separation as a function of period .
also shown is the size of the line - emitting region , @xmath36 , versus period .
both are calculated using the maximum black hole mass for each period assuming an 120 @xmath53 or 20 @xmath53 donor .
the orbital separation is greater than the upper limit of the size of the line - emitting region when the compact object mass is below 875 m@xmath3 for an 120 @xmath53 companion and below 128 @xmath53 for a 20 @xmath53 companion .
these masses are reduced if the inclination is lowered .
these results suggest that the most probable black hole mass is at most a factor of several above the usual stellar - mass black hole range .
@xcite proposed an orbital period of @xmath55 days for ngc 5408 x-1 , based on variations in the x - ray emission . with @xmath56 days , the mass function is @xmath57 13.4 m@xmath3 , implying a lower bound on the companion mass @xmath58 10.6 m@xmath3 .
it is interesting to determine if this period is consistent with other constraints on the system .
an orbital period of @xmath55 days would require a mean stellar density of @xmath59 g @xmath60 and , thus , a supergiant companion if mass transfer proceeds via roche - lobe overflow @xcite .
in particular , late f and early g supergiants have densities close to that required , although we caution that the high mass transfer rate needed to power the ulx may distort the spectral type of the star .
such stars have masses of 1012@xmath53 , consistent with the minimum mass derived from the mass function , and absolute magnitudes close to or below the upper limit quoted above .
the stellar radii are large , up to 0.7 au , but smaller than the orbital separation for this period and mass .
however , a companion mass so close to the lower bound on the mass function would require a very low mass compact object .
for a companion mass of 1012 @xmath53 , the compact object would have to be below 1 @xmath53 , which seems unlikely .
for a 5 @xmath53 black hole , one would need a donor of about 17 @xmath53 if the system is edge on . a higher black hole mass or a less extreme inclination would require an even higher mass companion .
these high companion masses contradict the required stellar density ; there is no star with both @xmath61 and @xmath62 g @xmath60 .
thus , either the orbital period is not near 115 days or mass transfer does not proceed via roche - lobe overflow .
we note that @xcite have suggested that 115 day periodicity may , instead , indicate a super - orbital period .
we thank manfred pakull and the referee for valuable comments .
this research received funding from the european community s seventh framework programme ( fp7/2007 - 2013 ) under grant agreement number itn 215212 `` black hole universe '' .
we acknowledge financial support from a marie curie reintegration grant under contract number perg05-ga-2009 - 243869 and from nasa grant nnx08aj26g._facilities : _ . | high - resolution optical spectra of the ultraluminous x - ray source ngc 5408 x-1 show a broad component with a width of @xmath0750 km / s in the heii and h@xmath1 lines in addition to the narrow component observed in these lines and [ o iii ] .
reanalysis of moderate - resolution spectra shows a similar broad component in the heii line .
the broad component likely originates in the ulx system itself , probably in the accretion disk .
the central wavelength of the broad heii line is shifted by @xmath2 km / s between the two observations .
if this shift represents motion of the compact object , then its mass is less than @xmath01800 m@xmath3 . |
it is usually not seriously discussed in normal textbooks on quantum mechanics about _ how to prepare an initial state_. it is , however , becoming an important subject not only from a view point of foundation of quantum mechanics , but also from a practical point of view , since we are rushing towards experimental realizations of the ideas for quantum information and computation @xcite . without establishing particular initial states assumed in several algorithms , we can not start any processes of the attractive ideas .
state preparation is one of the key elements to quantum information processing @xcite , and there are several theoretical proposals @xcite and experimental attempts @xcite . in the ideas for quantum information and computation , quantum states with high coherence , especially _ entangled states _ , play significant and essential roles . but
such `` clean '' states required for quantum information technologies are not easily found in nature , since many of them are fragile against environmental perturbations and suffer from _
decoherence_. therefore , there would often be a demand for preparing a desired _ pure state _ out of an arbitrary _ mixed state_. several schemes have been proposed for it , which are called `` purification , '' `` distillation , '' `` concentration , '' `` extraction , '' etc .
@xcite .
one of the simplest and easiest ways of state preparation is to resort to a projective measurement : a quantum system shall be in a pure state @xmath0 after it is measured and confirmed to be in the state @xmath0 .
such a strategy is not possible , however , in cases where the desired state @xmath0 can not be directly measured or where the relevant system is not available after the confirmation .
this is often the case for entangled states , which are the key resources to quantum information and computation .
this is why more elaborate purification protocols are required and several schemes of _ entanglement purification/ preparation _ have been proposed @xcite . recently , a novel mechanism to purify quantum states has been found and reported : _ purification through zeno - like measurements _ @xcite .
a pure state is extracted in a quantum system through a series of repeated measurements ( zeno - like measurements ) on another quantum system in interaction with the former .
since the relevant system to be purified is not directly measured in this scheme , it would be suitable for such situations mentioned above . in this article
, we discuss this scheme in detail and explore , on a heuristic basis , its potential as a useful and effective method of purification of qubits . the examples considered here are quite simple but still possess potential and practical applicability .
this article is organized as follows .
first , the basic framework of the purification is described in a general setting , and the conditions for the purification and its optimization are summarized in sec .
[ sec : framework ] , where some details which are not discussed in the first report @xcite are included .
it is then demonstrated in sec .
[ sec : single ] how it works and how it can be made optimal in a simplest example , i.e. , _ single - qubit purification _ , and a generalization to a multi - qubit case is considered in sec .
[ sec : initialization ] , which would afford us a useful method of _ initialization of multiple qubits_. one of the interesting applications of the present scheme is _ entanglement purification _ , which is discussed in sec .
[ sec : entanglementpurification ] and shown to be actually possible .
concluding remarks are given in sec .
[ sec : summary ] with some comments on possible extensions and future subjects .
appendices a e are supplied in order to demonstrate detailed calculations and proofs , that are not described in the text .
let us recapitulate the framework of the purification reported in @xcite .
we consider two quantum systems x and a interacting with each other ( fig .
[ fig : coupledsystem ] ) . the total system x+a is initially in a _ mixed _ state @xmath1 , from which we try to extract a pure state in a by controlling x. we first perform a measurement on x ( the zeroth measurement ) _ to confirm that it is in a state @xmath2_. if it is found in the state @xmath2 , the state of the total system is projected by the projection operator @xmath3 to yield @xmath4 where @xmath5 is the state of a after this zeroth confirmation and @xmath6 is the probability for this to happen .
we then let the total system start to evolve under a total hamiltonian @xmath7 and repeat the same measurement on x at regular time intervals @xmath8 .
after @xmath9 repetitions of successful confirmations , i.e. , after x is confirmed to be in the state @xmath2 _ successively _ @xmath9 times , the state of the total system , @xmath10 , is cast into the following form : [ eqn : state ] @xmath11\\ & = { { |{\phi}\rangle}_{\text{x}}\hspace*{-0.2truemm}{\langle{\phi}|}}\otimes \varrho_\text{a}^{(\tau)}(n ) , \label{eqn : statetotal}\displaybreak[0]\\ \varrho_\text{a}^{(\tau)}(n ) & = \bm{(}v_\phi(\tau)\bm{)}^n\varrho_\text{a } \bm{(}v_\phi^\dag(\tau)\bm{)}^n/\tilde{p}^{(\tau)}(n ) , \label{eqn : statea}\end{aligned}\ ] ] where @xmath12 , defined by @xmath13 is a projected time - evolution operator acting on the hilbert space of a , and @xmath14 is the normalization factor , @xmath15 \nonumber\displaybreak[0]\\ & = { \mathop{\text{tr}}\nolimits}_\text{a}[\bm{(}v_\phi(\tau)\bm{)}^n\varrho_\text{a } \bm{(}v_\phi^\dag(\tau)\bm{)}^n ] .
\label{eqn : yield}\end{aligned}\ ] ] note that we retain only those events where x is found in the state @xmath2 at _ every _ measurement ( including the zeroth one ) ; other events , resulting in failure to purify a , are discarded .
the normalization factor @xmath14 multiplied by @xmath16 , i.e. , @xmath17 , is nothing but the probability for the _ successful events _ and is the probability of obtaining the state given in ( [ eqn : state ] ) . for definiteness ,
let us restrict ourselves on finite - dimensional systems throughout this article and consider the spectral decomposition of the operator @xmath12 .
since the operator @xmath12 is not a hermitian operator , we should set up both right and left eigenvalue equations @xmath18 the eigenvalues @xmath19 are complex in general and bounded as @xmath20 ( see appendix [ app : bound ] ) .
here we assume for simplicity that the spectrum of the operator @xmath12 is not degenerate .
in such a case , the eigenvectors are orthogonal to each other in the sense @xmath21 and form a complete set in the hilbert space of system a , @xmath22 which readily leads to the spectral decomposition of the operator @xmath12 , @xmath23 ( in the following , we also normalize the right eigenvectors as @xmath24 . ) even in a general situation where the spectrum of the operator @xmath12 is degenerate , the diagonalization ( [ eqn : spectraldecomp ] ) is possible when and only when all the right eigenvectors @xmath25 are linearly independent of each other and form a complete basis @xcite . otherwise , the spectral decomposition is not like ( [ eqn : spectraldecomp ] ) , but in the `` jordan canonical form '' @xcite .
the diagonalizability of the operator @xmath12 is , however , not an essential assumption as clarified in appendix [ app : jordandecomposition ] .
it is now easy to observe the asymptotic behavior of the state of a , @xmath26 in ( [ eqn : statea ] ) .
since the eigenvalues @xmath19 are bounded like ( [ eqn : bound ] ) , each term in the expansion @xmath27 decays out and a single term dominates asymptotically as the number of measurements , @xmath9 , increases , @xmath28 _ provided _
\text{\textit{unique , discrete and non\-degenerate}}. \label{eqn : condition}\end{gathered}\ ] ] [ the word `` unique '' means that there is only one eigenvalue that has the maximum modulus and `` nondegenerate '' means that there is only one right eigenvector ( and a corresponding left eigenvector ) belonging to that maximal ( in magnitude ) eigenvalue . ]
thus , the state of a in ( [ eqn : statea ] ) approaches a pure state @xmath30 , @xmath31 this is the purification scheme proposed recently @xcite : extraction of a pure state @xmath30 through a series of repeated measurements on x. since we repeat measurements ( on x ) as in the case of the quantum zeno effect @xcite , we call such measurements `` zeno - like measurements '' @xcite . the final pure state @xmath30 is the eigenstate of the projected time - evolution operator @xmath12 belonging to the largest ( in magnitude ) eigenvalue @xmath32 and depends on the parameters @xmath8 , @xmath2 , and those in the hamiltonian @xmath7 .
it is , however , independent of the initial state @xmath1 .
the pure state @xmath33 is extracted from an _ arbitrary _ mixed state @xmath1 through the zeno - like measurements . by tuning such parameters mentioned above
, we have a possibility of extracting a desired pure state @xmath30 .
the above observation shows that the assumption of the diagonalizability in ( [ eqn : spectraldecomp ] ) is not essential but condition ( [ eqn : condition ] ) , i.e. , the existence of the _
unique , discrete and nondegenerate largest ( in magnitude ) eigenvalue @xmath32 _ , is crucial to the purification . for our purification mechanism to work ,
it is crucial that a single state is extracted and this is accomplished when these qualifications , i.e. , the uniqueness of the largest eigenvalue and the nondegeneracy of the eigenvector , are both met .
the diagonalizability of @xmath12 is not relevant to these conditions and is not essential to the purification .
this point is clarified in appendix [ app : jordandecomposition ] .
furthermore , note the asymptotic behavior of the success probability @xmath34 : it decays asymptotically as @xmath35 where @xmath36 stands for @xmath37 and @xmath38 .
the decay is governed by the eigenvalue @xmath32 , and therefore , an efficient purification is possible if @xmath32 satisfies the condition @xmath39 which suppresses the decay in ( [ eqn : decay ] ) to give the final ( nonvanishing ) success probability @xmath40 it is worth stressing that the condition ( [ eqn : optimizationi ] ) allows us to repeat the measurement as many times as we wish without running the risk of losing the success probability @xmath34 . in other words , high fidelity to the target state and nonvanishing success probability do not contradict each other in this scheme , but rather they can be achieved simultaneously . at the same time , if the other eigenvalues are much smaller than @xmath32 in magnitude , @xmath41 purification is achieved quickly . equations ( [ eqn : optimizationi ] ) and ( [ eqn : optimizationii ] ) are the conditions for the _ optimal purification _ , which we try to accomplish by adjusting parameters @xmath8 , @xmath2 , and those in the hamiltonian @xmath7 . in the following sections , we discuss the above purification scheme in more detail addressing a few specific examples , which are so simple but
still possess potential and practical applications in quantum information and computation .
let us first observe how the above mechanism works in the simplest example : we consider two qubits ( two two - level systems ) x and a interacting with each other , whose total hamiltonian is given by @xmath42 where @xmath43 are the pauli operators , @xmath44 are the ladder operators , and the frequencies @xmath45 and the coupling constant @xmath46 are real parameters .
we repeatedly confirm the state of x and purify qubit a , i.e. , we discuss a purification of a single qubit .
the four eigenvalues of the total hamiltonian @xmath7 in ( [ eqn : hamiltoniansingle ] ) are given by @xmath47\\ e^{(1)}_\pm&=(\omega_\text{x}+\omega_\text{a})/2\pm\delta , \displaybreak[0]\\ e^{(2)}&=\omega_\text{x}+\omega_\text{a},\end{aligned}\ ] ] and the corresponding eigenstates are [ eqn : eigenstatessingle ] @xmath48\\ { |{e^{(1)}_\pm}\rangle}_\text{xa } = { } & \frac{1}{\sqrt{2}}\,\biggl ( \epsilon(g)\sqrt{1\pm\frac{\omega_\text{x}-\omega_\text{a}}{2\delta}}{|{\uparrow\downarrow}\rangle}_\text{xa } \nonumber\displaybreak[0]\\ & \phantom{\frac{1}{\sqrt{2}}\,\biggl ( } { } \pm\sqrt{1\mp\frac{\omega_\text{x}-\omega_\text{a}}{2\delta } } { |{\downarrow\uparrow}\rangle}_\text{xa } \biggr),\displaybreak[0]\\ { |{e^{(2)}}\rangle}_\text{xa } = { } & { |{\uparrow\uparrow}\rangle}_\text{xa},\end{aligned}\ ] ] where @xmath49 @xmath50 is the sign function , and @xmath51 is the eigenstate of the operator @xmath52 belonging to the eigenvalue @xmath53 with the phase convention @xmath54 . hence , when the state of x , @xmath2 , is confirmed repeatedly at time intervals @xmath8 , the relevant operator to be investigated , the projected time - evolution operator @xmath12 , reads @xmath55\\ = { } & { { |{\uparrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\uparrow}| } } e^{-i(\omega_\text{x}+\omega_\text{a})\tau}\left [ \cos^2\!\frac{\theta}{2 } + e^{i(\omega_\text{x}+\omega_\text{a})\tau/2}\left ( \cos\delta\tau + i\frac{\omega_\text{x}-\omega_\text{a}}{2\delta}\sin\delta\tau \right)\sin^2\!\frac{\theta}{2 } \right]\nonumber\displaybreak[0]\\ & { } + { { |{\downarrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\downarrow}|}}\left [ \sin^2\!\frac{\theta}{2 } + e^{-i(\omega_\text{x}+\omega_\text{a})\tau/2}\left ( \cos\delta\tau -i\frac{\omega_\text{x}-\omega_\text{a}}{2\delta}\sin\delta\tau \right)\cos^2\!\frac{\theta}{2 } \right]\nonumber\displaybreak[0]\\ & { } -i\left ( { { |{\uparrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\downarrow}|}}e^{-i\varphi } + { { |{\downarrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\uparrow}|}}e^{i\varphi } \right)\frac{g}{\delta }
e^{-i(\omega_\text{x}+\omega_\text{a})\tau/2 } \sin\delta\tau\sin\frac{\theta}{2}\cos\frac{\theta}{2 } , \label{eqn : vsingle}\end{aligned}\ ] ] where the state @xmath2 is parameterized as @xmath56 and the set of angles @xmath57 characterizes the `` direction of ` spin ' x. '' if one of the two eigenvalues of the operator ( [ eqn : vsingle ] ) is larger in magnitude than the other , the condition for purification ( [ eqn : condition ] ) is fulfilled , and qubit a is purified into the eigenstate @xmath30 belonging to the larger ( in magnitude ) eigenvalue @xmath32 .
furthermore , if condition ( [ eqn : optimizationi ] ) , @xmath58 , is satisfied , we can purify with a nonvanishing success probability @xmath59 , and another condition ( [ eqn : optimizationii ] ) , @xmath60 , enables us to accomplish quick purification .
we try to achieve these conditions by tuning the parameters .
the first adjustment for the optimal purification is @xmath61 ( see appendix [ app : optimalityproof ] ) .
actually , if we choose @xmath62 , the eigenvalues of the projected time - evolution operator @xmath12 are given by @xmath63 and the eigenvectors belonging to them are @xmath64 it is clear that the magnitude of the eigenvalue @xmath32 is unity and that of @xmath65 , @xmath66 is less than unity provided @xmath67 both conditions ( [ eqn : condition ] ) and ( [ eqn : optimizationi ] ) are thus satisfied , and according to the theory presented in sec .
[ sec : framework ] , we have an optimal purification @xmath68 after the repeated confirmations of the state @xmath69 , qubit a is purified into @xmath70 with a _ nonvanishing _ probability @xmath71 .
similarly , another choice in ( [ eqn : optimizationsingle ] ) , i.e. , a series of repeated confirmations of the state @xmath72 , drives a into @xmath73 with a nonvanishing probability @xmath74 : @xmath75 the final success probability @xmath76 for the former choice @xmath62 or @xmath77 for the latter @xmath78 means that the target state @xmath79 or @xmath80 contained in the initial state @xmath1 is fully extracted . in this sense , the purification is optimal .
the second adjustment is for the fastest purification , which is realized by the condition @xmath81 at which @xmath82 in ( [ eqn : secondeigenvalue ] ) is the smallest : @xmath83 .
we can achieve it by tuning the time interval @xmath8 , for instance . and
success probability @xmath34 for single - qubit purification .
the pure state @xmath84 is extracted from the initial mixed state @xmath85 after repeated confirmations of the state @xmath86 .
parameters are @xmath87 , @xmath88 , @xmath89 for ( a ) and @xmath90 , @xmath91 for ( b ) , in the unit such that @xmath92 .
the time interval @xmath8 is tuned so as to satisfy the condition for the fastest purification ( [ eqn : fastestsingle ] ) in each case.,title="fig:",scaledwidth=45.0% ] + and success probability @xmath34 for single - qubit purification .
the pure state @xmath84 is extracted from the initial mixed state @xmath85 after repeated confirmations of the state @xmath86 .
parameters are @xmath87 , @xmath88 , @xmath89 for ( a ) and @xmath90 , @xmath91 for ( b ) , in the unit such that @xmath92 .
the time interval @xmath8 is tuned so as to satisfy the condition for the fastest purification ( [ eqn : fastestsingle ] ) in each case.,title="fig:",scaledwidth=45.0% ] to be more explicit , let us demonstrate the extraction of the pure state @xmath70 from the initial mixed state @xmath93 after x is confirmed to be in the state @xmath69 successfully @xmath9 times at time intervals @xmath8 , the state of qubit a and the probability for the successful confirmations read @xmath94^n { { |{\downarrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\downarrow}| } } } { 1+[1-(g/\delta)^2\sin^2\!\delta\tau]^n},\\ \displaystyle p^{(\tau)}(n ) = \frac{1}{2}\{1+[1-(g/\delta)^2\sin^2\!\delta\tau]^n\ } , \end{cases}\ ] ] respectively , which clearly confirm the limits ( [ eqn : toup ] ) unless @xmath95 ( @xmath96 ) , and the convergences are the fastest when the condition ( [ eqn : fastestsingle ] ) is satisfied .
( note that @xmath97 for the initial state considered here . ) in fig .
[ fig : fidelityyieldsingle](a ) , the success probability @xmath34 and the so - called fidelity to the target state @xmath98 , defined by @xmath99\\ & = { { } _ { \text{a}}\hspace*{-0.2truemm}\langle{\uparrow}|}\varrho_\text{a}^{(\tau)}(n ) { |{\uparrow}\rangle}_\text{a},\end{aligned}\ ] ] are shown as functions of the number of measurements , @xmath9 , for the initial state ( [ eqn : initialstatesingle ] ) , with the parameters @xmath87 , @xmath88 , @xmath92 , @xmath89 .
since the condition ( [ eqn : optimizationi ] ) , @xmath58 , is fulfilled , the decay of the success probability @xmath34 is suppressed to yield the finite value @xmath97 , and since the time interval @xmath8 is tuned so as to satisfy the condition for the fastest purification ( [ eqn : fastestsingle ] ) ( @xmath100 ) , the pure state @xmath70 is extracted after only @xmath101 or @xmath102 measurements . in an extreme case
where @xmath103 is possible , the extraction is achieved just after one measurement .
such a situation is depicted in fig .
[ fig : fidelityyieldsingle](b ) for the same initial state as in fig .
[ fig : fidelityyieldsingle](a ) with the parameter set @xmath90 , @xmath92 , @xmath104 .
the single - qubit purification in the previous section is too simple but is easily extended for multi - qubit cases . in the above example
, one may realize that the state @xmath79 is an eigenstate of the total hamiltonian ( [ eqn : hamiltoniansingle ] ) [ see ( [ eqn : eigenstatessingle ] ) ] and this is why the optimization condition ( [ eqn : optimizationi ] ) , @xmath58 , is achieved with @xmath62 and @xmath98 irrespectively of the choice of the time interval @xmath8 .
( the same argument applies to the case @xmath78 there . ) in the case of a multi - qubit system @xmath105 in fig .
[ fig : multiqubits ] , with nearest - neighbor interactions , @xmath106\\ & { } + g_\text{xa}(\sigma_+^\text{x}\sigma_-^\text{a } + \sigma_-^\text{x}\sigma_+^\text{a } ) + g_\text{ab}(\sigma_+^\text{a}\sigma_-^\text{b } + \sigma_-^\text{a}\sigma_+^\text{b})\nonumber\displaybreak[0]\\ & \phantom{{}+{}}{}+\cdots \label{eqn : hamiltonianmultiple}\end{aligned}\ ] ] ( @xmath107 ) , the state @xmath108 is an eigenstate of this total hamiltonian @xmath7 , and it is readily expected that the pure state @xmath109 is extracted by repeated projections onto the state @xmath69 , with the optimal success probability . similarly , repeated projections onto @xmath72 set every qubit into @xmath110 state , i.e. , into @xmath111 , optimally .
this would be useful for _ initialization of multiple qubits _ in a quantum computer . in order to make this idea more concrete ,
let us discuss in detail with a three - qubit system @xmath112 .
the important point is whether the condition for the purification ( [ eqn : condition ] ) is achievable , i.e. , whether all the eigenvalues @xmath19 except for the relevant one @xmath32 , associated with the eigenstate @xmath113 ( or @xmath114 ) , can actually be less than unity in magnitude . for simplicity , we consider the case where @xmath115 .
the eight eigenvalues of the total hamiltonian @xmath7 are given by [ eqn : eigenvaluesmultiple ] @xmath47\\ e^{(1)}_0&=\omega , & e^{(1)}_\pm&=\omega\pm\sqrt{2}\bar{g},\displaybreak[0]\\ e^{(2)}_0&=2\omega , & e^{(2)}_\pm&=2\omega\pm\sqrt{2}\bar{g},\displaybreak[0]\\ e^{(3)}&=3\omega,\end{aligned}\ ] ] and the corresponding eigenstates are [ eqn : eigenstatesmultiple ] @xmath116\\ { |{e^{(1)}_0}\rangle}_\text{xab } = { } & \cos\chi{|{\downarrow\downarrow\uparrow}\rangle}_\text{xab } -\sin\chi{|{\uparrow\downarrow\downarrow}\rangle}_\text{xab } , \displaybreak[0]\\ { |{e^{(1)}_\pm}\rangle}_\text{xab } = { } & \frac{1}{\sqrt{2 } } ( \sin\chi{|{\downarrow\downarrow\uparrow}\rangle}_\text{xab } + \cos\chi{|{\uparrow\downarrow\downarrow}\rangle}_\text{xab } \nonumber\displaybreak[0]\\ & \phantom{\frac{1}{\sqrt{2 } } ( } { } \pm{|{\downarrow\uparrow\downarrow}\rangle}_\text{xab } ) , \displaybreak[0]\\ { |{e^{(2)}_0}\rangle}_\text{xab } = { } & \cos\chi{|{\uparrow\uparrow\downarrow}\rangle}_\text{xab } -\sin\chi{|{\downarrow\uparrow\uparrow}\rangle}_\text{xab } , \displaybreak[0]\\ { |{e^{(2)}_\pm}\rangle}_\text{xab } = { } & \frac{1}{\sqrt{2 } } ( \sin\chi{|{\uparrow\uparrow\downarrow}\rangle}_\text{xab } + \cos\chi{|{\downarrow\uparrow\uparrow}\rangle}_\text{xab } \nonumber\displaybreak[0]\\ & \phantom{\frac{1}{\sqrt{2 } } ( } { } \pm{|{\uparrow\downarrow\uparrow}\rangle}_\text{xab } ) , \displaybreak[0]\\ { |{e^{(3)}}\rangle}_\text{xab } = { } & { |{\uparrow\uparrow\uparrow}\rangle}_\text{xab},\end{aligned}\ ] ] where @xmath117\\ \cos\chi=\frac{g_\text{xa}}{\sqrt{g_\text{xa}^2+g_\text{ab}^2 } } , \quad \sin\chi=\frac{g_\text{ab}}{\sqrt{g_\text{xa}^2+g_\text{ab}^2}}.\end{gathered}\ ] ] aiming at initializing qubits a and b into @xmath114 , we repeatedly project x onto the state @xmath72 at time intervals @xmath8 , and the relevant operator to be investigated reads @xmath118\\ = { } & { { |{\downarrow\downarrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\downarrow\downarrow}|}}\nonumber\displaybreak[0]\\ & { } + { { |{\uparrow\downarrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\uparrow\downarrow}| } } e^{-i\omega\tau}\cos\sqrt{2}\bar{g}\tau \nonumber\displaybreak[0]\\ & { } + { { |{\downarrow\uparrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\downarrow\uparrow}| } } e^{-i\omega\tau}(\cos^2\!\chi + \sin^2\!\chi\cos\sqrt{2}\bar{g}\tau ) \nonumber\displaybreak[0]\\ & { } -i{{|{\uparrow\downarrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\downarrow\uparrow}| } } e^{-i\omega\tau}\sin\chi\sin\sqrt{2}\bar{g}\tau \nonumber\displaybreak[0]\\ & { } -i{{|{\downarrow\uparrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\uparrow\downarrow}| } } e^{-i\omega\tau}\sin\chi\sin\sqrt{2}\bar{g}\tau \nonumber\displaybreak[0]\\ & { } + { { |{\uparrow\uparrow}\rangle}_{\text{ab}}\hspace*{-0.2truemm}{\langle{\uparrow\uparrow}| } } e^{-2i\omega\tau}(\sin^2\!\chi + \cos^2\!\chi\cos\sqrt{2}\bar{g}\tau).\end{aligned}\ ] ] the target state @xmath114 is an eigenstate of this operator belonging to the eigenvalue @xmath119 , which satisfies the optimization condition ( [ eqn : optimizationi ] ) , and the other three eigenvalues are give by [ eqn : initialize2eigenvalues ] @xmath120\\ & { } \mp\sin\frac{\bar{g}\tau}{\sqrt{2 } } \sqrt{\cos^4\!\chi\sin^2\!\frac{\bar{g}\tau}{\sqrt{2 } } -4\sin^2\!\chi\cos^2\!\frac{\bar{g}\tau}{\sqrt{2 } } } \biggr),\displaybreak[0]\\ \lambda_{\uparrow\uparrow } = { } & e^{-2i\omega\tau}\left ( 1 - 2\cos^2\!\chi\sin^2\!\frac{\bar{g}\tau}{\sqrt{2 } } \right ) .
\label{eqn : initialize2eigenvaluesupup}\end{aligned}\ ] ] if these three eigenvalues are all less than unity in magnitude , the condition for the purification ( [ eqn : condition ] ) is satisfied , and the initialized state @xmath114 is extracted from an arbitrary mixed state @xmath1 , with a nonvanishing success probability @xmath121 .
( note that the left eigenvector belonging to the eigenvalue @xmath122 is @xmath123 . )
such a situation is realized provided @xmath124 which is clearly seen from fig .
[ fig : eigenvaluesmultiple ] and a proof in appendix [ app : eigenvaluesmultiple ] .
( solid line ) , @xmath125 ( dashed line ) , and @xmath126 ( dotted line ) in ( [ eqn : initialize2eigenvalues ] ) , as functions of @xmath127 . in this figure
, we set @xmath128 .
note that @xmath129 within each range @xmath130 ( @xmath131 ) , where @xmath132 ( @xmath133 ) is defined by @xmath134 , and @xmath135 when @xmath128.,scaledwidth=45.0% ] and success probability @xmath34 for two - qubit initialization . through the repeated confirmations of the state @xmath72 ,
qubits a and b are initialized into @xmath114 from the thermal equilibrium state of the total system at temperature @xmath136 , i.e. , @xmath137 with @xmath138 .
parameters are @xmath139 , @xmath140 , @xmath141 , @xmath142 .
the time interval @xmath8 is tuned so as to make @xmath143 the smallest , which is for the fastest initialization ( see fig . [
fig : eigenvaluesmultiple]).,scaledwidth=45.0% ] the final success probability @xmath121 is again optimal , in the sense that the target state @xmath144 contained in the initial state @xmath1 is fully extracted .
the above argument reveals the possibility of initialization at least for two qubits .
initialization of two qubits into @xmath114 from the thermal equilibrium state of the total system at temperature @xmath136 , i.e. , @xmath137 with @xmath138 , is demonstrated in fig .
[ fig : initialize2 ] . note that it is effective when @xmath145 , since in such a case , @xmath144 is the ground state of the total system .
the analytic formula for the final success probability is @xmath146^{-1}$ ] .
it is natural to expect that the same mechanism also works for systems with more qubits as in fig .
[ fig : multiqubits ] .
it is hard to imagine that the magnitudes of eigenvalues of @xmath147 other than the relevant one @xmath148 ( whose magnitude is unity ) is also unity _ irrespectively of the values of parameters_. further detailed investigations on its efficiency , robustness , and so on , will certainly clarify the possibility of a new useful procedure for initializing multiple qubits .
one of the most significant issues in the field of quantum information and computation is how to prepare _ entanglement _ , and therefore , it is interesting and important to examine whether the present scheme can realize _ entanglement purification / preparation_. we show , in this section , that it is actually possible . in order to demonstrate it explicitly ,
let us discuss a simple hamiltonian @xmath149\\ & { } + g(\sigma_+^\text{x}\sigma_-^\text{a } + \sigma_-^\text{x}\sigma_+^\text{a } ) + g(\sigma_+^\text{x}\sigma_-^\text{b } + \sigma_-^\text{x}\sigma_+^\text{b } ) .
\label{eqn : hamiltonianentanglement}\end{aligned}\ ] ] the control qubit x is coupled to qubits a and b as in fig .
[ fig : threequbits ] .
we confirm x to be in the state @xmath2 repeatedly at time intervals @xmath8 and end up with an extraction of an entanglement between a and b , which are initially in a mixed state @xmath1 .
, in qubits a and b.,scaledwidth=30.0% ] the spectrum of the total hamiltonian @xmath7 is already given in ( [ eqn : eigenvaluesmultiple ] ) with @xmath150 replaced by @xmath151 , and the eigenstates are [ eqn : eigenstatesentanglement ] @xmath116\\ { |{e^{(1)}_0}\rangle}_\text{xab } = { } & { |{{\downarrow}\psi^-}\rangle}_\text{xab},\displaybreak[0]\\ { |{e^{(1)}_\pm}\rangle}_\text{xab } = { } & \frac{1}{\sqrt{2}}[{|{{\downarrow}\psi^+}\rangle}_\text{xab } \pm\epsilon(g){|{\uparrow\downarrow\downarrow}\rangle}_\text{xab } ] , \displaybreak[0]\\ { |{e^{(2)}_0}\rangle}_\text{xab } = { } & { |{{\uparrow}\psi^-}\rangle}_\text{xab},\displaybreak[0]\\ { |{e^{(2)}_\pm}\rangle}_\text{xab } = { } & \frac{1}{\sqrt{2}}[{|{{\uparrow}\psi^+}\rangle}_\text{xab } \pm\epsilon(g){|{\downarrow\uparrow\uparrow}\rangle}_\text{xab } ] , \displaybreak[0]\\ { |{e^{(3)}}\rangle}_\text{xab } = { } & { |{\uparrow\uparrow\uparrow}\rangle}_\text{xab}.\end{aligned}\ ] ] the relevant projected time - evolution operator is given , in this case , by @xmath152\\ = { } & { { |{\psi^-}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\psi^-}| } } e^{-i\omega\tau}\left ( \sin^2\!\frac{\theta}{2}+e^{-i\omega\tau}\cos^2\!\frac{\theta}{2 } \right)\nonumber\displaybreak[0]\\ & { } + { { |{\downarrow\downarrow}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\downarrow\downarrow}|}}\left ( \sin^2\!\frac{\theta}{2 } + e^{-i\omega\tau}\cos\sqrt{2}g\tau\cos^2\!\frac{\theta}{2 } \right)\nonumber\displaybreak[0]\\ & { } + { { |{\psi^+}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\psi^+}| } } e^{-i\omega\tau}\cos\sqrt{2}g\tau\left ( \sin^2\!\frac{\theta}{2}+e^{-i\omega\tau}\cos^2\!\frac{\theta}{2 } \right)\nonumber\displaybreak[0]\\ & { } + { { |{\uparrow\uparrow}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\uparrow\uparrow}| } } e^{-2i\omega\tau}\left ( \cos\sqrt{2}g\tau\sin^2\!\frac{\theta}{2 } + e^{-i\omega\tau}\cos^2\!\frac{\theta}{2 } \right)\nonumber\displaybreak[0]\\ & { } -i\left ( { { |{\downarrow\downarrow}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\psi^+}|}}e^{i\varphi } + { { |{\psi^+}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\downarrow\downarrow}|}}e^{-i\varphi } \right)e^{-i\omega\tau}\sin\sqrt{2}g\tau\sin\frac{\theta}{2 } \cos\frac{\theta}{2}\nonumber\displaybreak[0]\\ & { } -i\left ( { { |{\uparrow\uparrow}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\psi^+}|}}e^{-i\varphi } + { { |{\psi^+}\rangle}_{\mathrm{ab}}\hspace*{-0.2truemm}{\langle{\uparrow\uparrow}|}}e^{i\varphi } \right)e^{-2i\omega\tau}\sin\sqrt{2}g\tau\sin\frac{\theta}{2 } \cos\frac{\theta}{2},\end{aligned}\ ] ] where @xmath153 are the two of the four bell states @xmath154 , @xmath155 , and @xmath2 is parameterized as in ( [ eqn : spinparametrization ] ) . since the hamiltonian ( [ eqn : hamiltonianentanglement ] ) is symmetric under the exchange between a and b , @xmath12 splits into two sectors : the singlet sector and the triplet one .
the singlet state @xmath156 is apparently one of the four eigenstates of @xmath12 belonging to the eigenvalue @xmath157 and hence , we can extract an entangled state , i.e. , the bell state @xmath158 , after a number of measurements on x , provided ( i ) the eigenvalue @xmath159 is larger in magnitude than any other eigenvalues .
furthermore , if ( ii ) condition ( [ eqn : optimizationi ] ) , i.e. , @xmath160 , is achieved , @xmath158 is extracted with an optimal probability @xmath161 , which is again optimal in the same sense as in the preceding examples , i.e. , the target entangled state @xmath158 contained in the initial state @xmath1 has been fully extracted .
requirement ( ii ) is fulfilled by the choice of the parameters as @xmath162 ( @xmath131 ) or @xmath163 , but the latter choice violates requirement ( i ) .
it is , therefore , necessary that [ eqn : conditionentanglement ] @xmath164 ( note that the first condition is automatically satisfied without tuning the time interval @xmath8 , when @xmath165 . )
furthermore , one can prove as in appendix [ app : eigenvaluesentanglement ] that requirement ( i ) is met , under the condition ( [ eqn : conditionentanglementi ] ) , provided @xmath166 the existence of such a parameter set satisfying ( [ eqn : conditionentanglement ] ) explicitly discloses the possibility of extracting entanglement through zeno - like measurements . in the case of the choice @xmath167 for example
, the four eigenvalues are given by @xmath168\\ \lambda_\pm&=1-\frac{1}{2}\sin\frac{g\tau}{\sqrt{2}}\left ( 3\sin\frac{g\tau}{\sqrt{2 } } \pm\epsilon(g)\sqrt{1 - 9\cos^2\!\frac{g\tau}{\sqrt{2 } } } \right),\end{aligned}\ ] ] whose magnitudes behave as in fig .
[ fig : eigenvaluesmultiple ] but with @xmath126 replaced by @xmath169 . and success probability @xmath34 for entanglement purification
. the entangled state @xmath158 is extracted from ( a ) a product state @xmath170 and ( b ) the thermal state @xmath137 at temperature @xmath171 , through repeated confirmations of the state @xmath172 .
parameters are @xmath165 , @xmath173 for ( a ) , and @xmath165 , @xmath141 , @xmath174 for ( b ) , in the unit such that @xmath92 , where @xmath132 is defined in the caption of fig .
[ fig : eigenvaluesmultiple ] . for the initial thermal state in
( b ) with @xmath165 , the success probability for the zeroth confirmation is given by @xmath175 for any set of parameters @xmath176 , and the final value @xmath177^{-1}$ ] becomes largest at @xmath178.,title="fig:",scaledwidth=45.0% ] + and success probability @xmath34 for entanglement purification .
the entangled state @xmath158 is extracted from ( a ) a product state @xmath170 and ( b ) the thermal state @xmath137 at temperature @xmath171 , through repeated confirmations of the state @xmath172 .
parameters are @xmath165 , @xmath173 for ( a ) , and @xmath165 , @xmath141 , @xmath174 for ( b ) , in the unit such that @xmath92 , where @xmath132 is defined in the caption of fig .
[ fig : eigenvaluesmultiple ] . for the initial thermal state in
( b ) with @xmath165 , the success probability for the zeroth confirmation is given by @xmath175 for any set of parameters @xmath176 , and the final value @xmath177^{-1}$ ] becomes largest at @xmath178.,title="fig:",scaledwidth=45.0% ] the extraction of the entangled state @xmath158 is demonstrated in fig .
[ fig : entanglement ] , from a product state @xmath170 and from the thermal equilibrium state @xmath179 at temperature @xmath171 .
the examples presented in this article demonstrate how the present purification scheme works , and suggest a few potential applications , even though the analyses are heuristically based and no general `` optimization '' theory or strategy has been given .
remarkable features of the scheme are summarized as follows .
( i ) the first point is the simplicity .
many of the other proposed procedures are composed of several steps with different operations , such as rotation , cnot operation , and measurement @xcite . in the present scheme , on the other hand , one has only to repeat one and the same measurement .
( ii ) furthermore , the `` optimal '' success probability is possible in the sense that the target state contained in the initial state is fully extracted . in several other methods @xcite , on the contrary , it decays to zero as the fidelity approaches unity @xcite .
( iii ) the number of measurements required for purification is considerably reduced by appropriate choices of parameters , and purification is attainable after only a few steps .
another point to be stressed is the flexibility .
while many of the other schemes @xcite are designed for specific systems , the framework is presented in sec .
[ sec : framework ] on a general setting , and there are diverse systems and purposes which fit the present scheme .
we have already observed , in this article , two different applications on the same idea : initialization and entanglement purification .
additional ideas or slight modifications to the basic scheme would provide us with various methods of state preparation .
an interesting extension of the present scheme is for extraction of entanglement _ between two spatially - separated qubits _
@xcite , which is often necessary for quantum communication , quantum teleportation , and so on .
( the original protocols for entanglement purification @xcite are aimed at this purpose . )
it is actually possible and will be reported elsewhere @xcite .
one of the other possible extensions is to go beyond a method of extracting quantum state .
it would be interesting , for example , if we could find a novel method of _ transferring _ quantum state @xcite rather than extracting it . in this article , only qubit systems , i.e. , finite - dimensional systems ,
have been discussed .
one has to keep in mind that the condition ( [ eqn : condition ] ) plays a crucial role in the present purification scheme .
if this condition is met , however , it works for infinite - dimensional ones as well .
in fact , a harmonic oscillator , which has an infinite number of energy levels , can be purified through the present method , which is explicitly demonstrated in @xcite .
this also shows the broad range of applicability of the scheme .
it is not obvious , however , whether one can purify systems with continuous spectra , since they seem , at first sight , unlikely to satisfy the condition for purification ( [ eqn : condition ] ) , especially the discreteness of the eigenvalues .
this point is one of the interesting future subjects , since it would be required in some cases to purify quantum states in the presence of environmental systems , namely , under dissipation and/or dephasing .
the simplicity and the efficiency mentioned above would facilitate practical experimental applications of the present scheme .
the flexibility allows one to apply it to various kinds of systems intended for quantum information and computation , such as optical setups @xcite , ion - trap systems @xcite , solid - state quantum computers @xcite , and so on . in practice , one should face many unwanted factors , and robustness of the method against them is crucial . in the present scheme ,
it is often required to tune certain parameters in order to extract a desired pure state , and it is an important subject to clarify how precise the tuning should be and how much error the method suffers from when the parameters are mistuned .
it is also a remained issue to explore how ideal projective measurements are realized in actual experiments .
investigations on these points are now in progress . the authors acknowledge useful and helpful discussions with professor i. ohba .
this work is partly supported by a grant for the 21st century coe program ( physics of self - organization systems ) at waseda university and a grant - in - aid for priority areas research ( b ) ( no .
13135221 ) from the ministry of education , culture , sports , science and technology , japan , by a grant - in - aid for scientific research ( c ) ( no .
14540280 ) from the japan society for the promotion of science , by a waseda university grant for special research projects ( no .
2002a-567 ) , and by the bilateral italian - japanese project 15c1 on `` quantum information and computation '' of the italian ministry for foreign affairs .
let us prove that the eigenvalues @xmath19 of the projected time - evolution operator @xmath12 are bounded as in ( [ eqn : bound ] ) . for an arbitrary state of a , say @xmath180 , @xmath181\\ & = \|{{}_{\text{x}}\hspace*{-0.2truemm}\langle{\phi}|}e^{-ih_\text{tot}\tau } { |{\phi}\rangle}_\text{x}{|{\psi}\rangle}_\text{a}\|^2 \nonumber\displaybreak[0]\\ & \le\|e^{-ih_\text{tot}\tau}{|{\phi}\rangle}_\text{x } { |{\psi}\rangle}_\text{a}\|^2 \nonumber\displaybreak[0]\\ & = 1 .
\label{eqn : inequality}\end{aligned}\ ] ] hence , by setting @xmath182 [ a right eigenvector of the operator @xmath12 ] and noting @xmath184 , we obtain the inequality ( [ eqn : bound ] ) .
as is clear from this proof , the bound ( [ eqn : bound ] ) reflects unitarity of the time - evolution operator @xmath185 .
it is assumed in sec . [
sec : framework ] that the projected time - evolution operator @xmath12 is diagonalized like ( [ eqn : spectraldecomp ] ) , but it is not the case if some of its eigenvalues are degenerated .
here we show , however , that the assumption of the diagonalizability is not essential to the purification .
when an eigenvalue @xmath19 of the ( finite - dimensional ) operator @xmath12 is @xmath186-fold degenerate , there do not always exist @xmath186 linearly independent eigenvectors .
this fact spoils the diagonalizability of the operator @xmath12 .
there exist @xmath187 linearly independent right eigenvectors @xmath188 ( @xmath189 ) belonging to the eigenvalue @xmath19 ( @xmath190 is called `` dimension of the eigenspace '' ) , and one can find @xmath191 linearly independent `` generalized eigenvectors '' @xmath188 ( @xmath192 ) which are subjected to the conditions @xmath193 and linearly independent of the eigenvectors @xmath188 ( @xmath189 ) @xcite .
the right vectors @xmath188 ( @xmath194 ) then form a complete set within the subspace associated with the eigenvalue @xmath19 , and there exist corresponding left vectors @xmath195 ( @xmath194 ) , which satisfy the orthonormality @xmath196 and completeness conditions @xmath197
the operator @xmath12 is now expanded as [ eqn : jordan ] @xmath198 with @xmath199 which is the most general form of spectral decomposition and is called `` jordan canonical form '' @xcite . note the relations @xmath200\\ \mathcal{d}_m\mathcal{p}_n&=\mathcal{p}_n\mathcal{d}_m = \mathcal{d}_m\delta_{mn},\displaybreak[0]\\ \mathcal{d}_m\mathcal{d}_n&=0\quad(m\neq n),\end{aligned}\ ] ] and @xmath201 from the spectral decomposition ( [ eqn : jordan ] ) , it is easily deduced that @xmath202 where @xmath203 therefore , _ if the largest ( in magnitude ) eigenvalue is unique _ , which is denoted by @xmath32 , _ and nondegenerate _
( i.e. , @xmath204 , @xmath205 , @xmath206 ) , the single term in the expansion ( [ eqn : generalizedexpansion ] ) again dominates asymptotically like ( [ eqn : mechanism ] ) ( note that @xmath207 for large @xmath9 ) , which leads to the same conclusion as ( [ eqn : purification ] )
. the purification does not suffer from degeneracy in the other eigenvalues than the largest ( in magnitude ) one @xmath32 .
the crucial condition to the purification is ( [ eqn : condition ] ) .
we show here that the condition ( [ eqn : optimizationsingle ] ) together with ( [ eqn : optimizationsingleii ] ) is the necessary and sufficient condition for the optimal purification with both ( [ eqn : condition ] ) and ( [ eqn : optimizationi ] ) for model ( [ eqn : hamiltoniansingle ] ) .
first , we try to achieve the upper bound in the inequality ( [ eqn : inequality ] ) , i.e. , @xmath208 , in model ( [ eqn : hamiltoniansingle ] ) . if such a state @xmath180 is found and is an eigenstate of the operator @xmath12 , say @xmath25 , we have @xmath209 . as is easily seen from ( [ eqn : inequality ] ) , the equality holds only when @xmath210 is satisfied , where @xmath211 is a vector perpendicular to @xmath2 in ( [ eqn : spinparametrization ] ) , i.e. , @xmath212 equation ( [ eqn : equalitycondition ] ) means that the operator @xmath213 should have a zero eigenvalue , and hence @xmath214 for model ( [ eqn : hamiltoniansingle ] ) , the operator @xmath215 reads @xmath216\sin\frac{\theta}{2}\cos\frac{\theta}{2 } \nonumber\displaybreak[0]\\ & { } -{{|{\downarrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\downarrow}|}}\left [ 1-e^{-i(\omega_\text{x}+\omega_\text{a})\tau/2}\left ( \cos\delta\tau -i\frac{\omega_\text{x}-\omega_\text{a}}{2\delta}\sin\delta\tau \right ) \right]\sin\frac{\theta}{2}\cos\frac{\theta}{2 } \nonumber\displaybreak[0]\\ & { } + i{{|{\uparrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\downarrow}| } } \frac{g}{\delta}e^{-i\varphi } e^{-i(\omega_\text{x}+\omega_\text{a})\tau/2 } \sin\delta\tau\cos^2\!\frac{\theta}{2 } -i{{|{\downarrow}\rangle}_{\text{a}}\hspace*{-0.2truemm}{\langle{\uparrow}| } } \frac{g}{\delta}e^{i\varphi } e^{-i(\omega_\text{x}+\omega_\text{a})\tau/2 } \sin\delta\tau\sin^2\!\frac{\theta}{2},\end{aligned}\ ] ] and @xmath217\sin^2\!\theta.\ ] ] condition ( [ eqn : equalitycondition ] ) , namely ( [ eqn : equalitycondition2 ] ) , is hence reduced to @xmath218 or @xmath219 in the first case ( [ eqn : firstpossibility ] ) , both conditions ( [ eqn : condition ] ) and ( [ eqn : optimizationi ] ) are satisfied unless @xmath95 ( @xmath96 ) as is explained around ( [ eqn : optimizationsingle])([eqn : optimizationsingleii ] ) . in the second case ( [ eqn : secondpossibility ] ) , on the other hand , the projected time - evolution operator reads @xmath220 and the eigenvalue @xmath221 is degenerated , i.e. , condition ( [ eqn : condition ] ) is not fulfilled .
therefore , the necessary and sufficient condition for the optimal purification in model ( [ eqn : hamiltoniansingle ] ) is given by the first choice ( [ eqn : firstpossibility ] ) [ i.e. , ( [ eqn : optimizationsingle ] ) ] with ( [ eqn : optimizationsingleii ] ) .
we here outline the proof of the necessary and sufficient condition for the optimal two - qubit initialization , eq .
( [ eqn : initializationcondition ] ) , in sec .
[ sec : initialization ] . what we have to show is how to make the eigenvalues @xmath222 and @xmath126 in ( [ eqn : initialize2eigenvalues ] ) all less than unity in magnitude .
the eigenvalues @xmath222 are the solutions to an eigenvalue equation @xmath223\\ { } + 1 - 2\cos^2\!\chi\sin^2\!\frac{\bar{g}\tau}{\sqrt{2}}&=0 .
\label{eqn : eigenvalueeqmultiple}\end{aligned}\ ] ] we clarify when this equation has a solution whose magnitude is unity .
seeking such a solution , we insert @xmath224 into ( [ eqn : eigenvalueeqmultiple ] ) to obtain the conditions @xmath225\\ \cos\theta\left ( \cos\theta -\cos^2\!\frac{\bar{g}\tau}{\sqrt{2 } } + \sin^2\!\chi\sin^2\!\frac{\bar{g}\tau}{\sqrt{2 } } \right)&\nonumber\displaybreak[0]\\ { } -\cos^2\!\chi\sin^2\!\frac{\bar{g}\tau}{\sqrt{2}}&=0,\end{aligned}\ ] ] which are reduced to @xmath226\\ \intertext{or } \cos\frac{\bar{g}\tau}{\sqrt{2}}=0\quad&\text{and}\quad \cos\theta=-1.\end{aligned}\ ] ] ( note that @xmath227 and @xmath228 , since it is assumed that @xmath229 . ) it is easy to see from ( [ eqn : initialize2eigenvaluesupup ] ) that we have @xmath230 when @xmath231 , and in summary , the magnitude of one of the eigenvalues @xmath222 and @xmath126 becomes unity only when @xmath232 the condition for the initialization in sec .
[ sec : initialization ] is thus proved to be ( [ eqn : initializationcondition ] ) .
see also fig .
[ fig : eigenvaluesmultiple ] .
the necessary and sufficient condition ( [ eqn : conditionentanglement ] ) for the entanglement purification in sec .
[ sec : entanglementpurification ] is proved in a similar manner to that in appendix [ app : eigenvaluesmultiple ] .
the eigenvalues @xmath222 and @xmath169 under the condition ( [ eqn : conditionentanglementi ] ) are the solutions to an eigenvalue equation @xmath233\\ { } + 2\sin^2\!\frac{\theta}{2}\cos^2\!\frac{\theta}{2 } \cos^2\!\frac{g\tau}{\sqrt{2}}\left ( \kappa-\frac{1}{2}\sin\frac{g\tau}{\sqrt{2 } } \right)=0\end{aligned}\ ] ] with @xmath234 $ ] .
seeking a solution @xmath235 with unit magnitude , we insert @xmath236 into this equation to obtain @xmath237\\ & \phantom{\sin\theta\,\biggl [ } { } -\left(2-\frac{1}{2}\sin^2\!\theta\right ) \sin^2\!\frac{g\tau}{\sqrt{2}}\left ( 2-\sin^2\!\frac{g\tau}{\sqrt{2 } } \right ) \biggr]=0,\displaybreak[0]\\ & \cos\theta\,\biggl [ 2\cos^2\!\theta
-\left(3 - 4\sin^2\!\frac{g\tau}{\sqrt{2}}\right)\cos\theta \nonumber\displaybreak[0]\\ & \phantom{\sin\theta\,\biggl [ } { } -\left(2-\frac{1}{2}\sin^2\!\theta\right ) \sin^2\!\frac{g\tau}{\sqrt{2}}\left ( 2-\sin^2\!\frac{g\tau}{\sqrt{2 } } \right ) \biggr]\nonumber\displaybreak[0]\\ & \quad\phantom{= { } } { } + 1 - 2\sin^4\!\frac{g\tau}{\sqrt{2 } } -\frac{1}{2}\sin^2\!\theta\sin^2\!\frac{g\tau}{\sqrt{2}}\left ( 2 - 3\sin^2\!\frac{g\tau}{\sqrt{2 } } \right)\nonumber\displaybreak[0]\\ & \quad=0,\end{aligned}\ ] ] extraction of entanglement is not possible when ( [ eqn : entanglementnogoi ] ) or ( [ eqn : entanglementnogoii ] ) is satisfied , and therefore , the condition for the entanglement purification in sec .
[ sec : entanglementpurification ] is given by ( [ eqn : conditionentanglement ] ) .
k. vogel , v. m. akulin , and w. p. schleich , phys .
. lett . * 71 * , 1816 ( 1993 ) ; a. s. parkins , p. marte , p. zoller , and h. j. kimble , _ ibid . _ * 71 * , 3095 ( 1993 ) ; b. m. garraway , b. sherman , h. moya - cessa , p. l. knight , and g. kurizki , phys .
a * 49 * , 535 ( 1994 ) ; c. k. law and j. h. eberly , phys .
rev . lett . * 76 * , 1055 ( 1996 ) ; b. kneer and c. k. law , phys .
a * 57 * , 2096 ( 1998 ) , and references therein .
j. i. cirac and p. zoller , phys .
a * 50 * , 2799(r ) ( 1994 ) ; m. freyberger , p. k. aravind , m. a. horne , and a. shimony , _ ibid . _ * 53 * , 1232 ( 1996 ) ; m. b. plenio , s. f. huelga , a. beige , and p. l. knight , _ ibid .
_ * 59 * , 2468 ( 1999 ) ; j. hong and h .- w .
lee , phys .
lett . * 89 * , 237901 ( 2002 ) ; c. marr , a. beige , and g. rempe , phys .
a * 68 * , 033817 ( 2003 ) . c. cabrillo
, j. i. cirac , p. garca - fernndez , and p. zoller , phys .
a * 59 * , 1025 ( 1999 ) ; l .- m .
duan , m. d. lukin , j. i. cirac , and p. zoller , nature ( london ) * 414 * , 413 ( 2001 ) ; a. messina , eur .
j. d * 18 * , 379 ( 2002 ) ; d. e. browne and m. b. plenio , phys .
a * 67 * , 012325 ( 2003 ) ; x .-
feng , z .- m .
zhang , x .- d .
li , s .- q .
gong , and z .- z .
xu , phys .
lett . * 90 * , 217902 ( 2003 ) ; l .- m . duan and h. j. kimble , _ ibid . _ * 90 * , 253601 ( 2003 ) ; d. e. browne , m. b. plenio , and s. f. huelga , _ ibid . _ * 91 * , 067901 ( 2003 ) .
b. demarco , a. ben - kish , d. leibfried , v. meyer , m. rowe , b. m. jelenkovi , w. m. itano , j. britton , c. langer , t. rosenband , and d. j. wineland , phys .
lett . * 89 * , 267901 ( 2002 ) ; a. ben - kish , b. demarco , v. meyer , m. rowe , j. britton , w. m. itano , b. m. jelenkovi , c. langer , d. leibfried , t. rosenband , and d. j. wineland , _ ibid . _
* 90 * , 037902 ( 2003 ) , and references therein .
f. schmidt - kaler , h. hffner , m. riebe , s. gulde , g. p. t. lancaster , t. deuschle , c. becher , c. f. roos , j. eschner , and r. blatt , nature ( london ) * 422 * , 408 ( 2003 ) ; f. schmidt - kaler , h. hffner , s. gulde , m. riebe , g. p. t. lancaster , t. deuschle , c. becher , w. hnsel , j. eschner , c. f. roos , and r. blatt , appl .
b * 77 * , 789 ( 2003 ) , and references therein .
y. nakamura , yu . a. pashkin , and j. s. tsai , nature ( london ) * 398 * , 786 ( 1999 ) ; t. yamamoto , yu .
a. pashkin , o. astafiev , y. nakamura , and j. s. tsai , _ ibid .
_ * 425 * , 941 ( 2003 ) , and references therein . c. h. bennett , g. brassard , s. popescu , b. schumacher , j. a. smolin , and w. k. wootters , phys .
lett . * 76 * , 722 ( 1996 ) ; * 78 * , 2031(e ) ( 1997 ) ; c. h. bennett , d. p. divincenzo , j. a. smolin , and w. k. wootters , phys .
a * 54 * , 3824 ( 1996 ) .
t. yamamoto , m. koashi , .
k. zdemir , and n. imoto , nature ( london ) * 421 * , 343 ( 2003 ) ; z. zhao , t. yang , y .- a .
chen , a .- n .
zhang , and j .- w .
pan , phys .
rev . lett . *
90 * , 207901 ( 2003 ) ; a. vaziri , j .- w .
pan , t. jennewein , g. weihs , and a. zeilinger , _ ibid . _
* 91 * , 227902 ( 2003 ) . for reviews , see h. nakazato , m. namiki , and s. pascazio , int . j. mod .
b * 10 * , 247 ( 1996 ) ; d. home and m. a. b. whitaker , ann .
* 258 * , 237 ( 1997 ) ; p. facchi and s. pascazio , in _ progress in optics _ , edited by e. wolf ( elsevier , amsterdam , 2001 ) , vol .
42 , p. 147 .
it should be noted , however , that the time interval @xmath8 in this scheme is not necessarily small as in the ordinary zeno measurements , and the purification ( [ eqn : purification ] ) is not due to the quantum zeno effect .
if the ordinary zeno limit @xmath239 and @xmath240 ( @xmath241 fixed ) @xcite is taken in the present scheme , a quantum zeno effect appears yielding the so - called `` quantum zeno dynamics '' @xcite , which is unitary and provides us with a quite different effect from the one discussed in this article .
p. facchi , a. g. klein , s. pascazio , and l. s. schulman , phys .
a * 257 * , 232 ( 1999 ) ; p. facchi , v. gorini , g. marmo , s. pascazio , and e. c. g. sudarshan , _ ibid . _ * 275 * , 12 ( 2000 ) ; p. facchi , s. pascazio , a. scardicchio , and l. s. schulman , phys .
a * 65 * , 012108 ( 2001 ) ; p. facchi and s. pascazio , phys .
lett . * 89 * , 080401 ( 2002 ) .
s. bose , phys .
* 91 * , 207901 ( 2003 ) ; l. amico , a. osterloh , f. plastina , r. fazio , and g. m. palma , phys .
a * 69 * , 022304 ( 2004 ) ; v. subrahmanyam , _ ibid . _ * 69 * , 034304 ( 2004 ) ; m. christandl , n. datta , a. ekert , and a. j. landahl , quant - ph/0309131 ( 2003 ) . | a novel method of purification , _ purification through zeno - like measurements _ [ h. nakazato , t. takazawa , and k. yuasa , phys . rev .
lett . * 90 * , 060401 ( 2003 ) ] , is discussed extensively and applied to a few simple qubit systems .
it is explicitly demonstrated how it works and how it is optimized . as possible applications , schemes for _ initialization of multiple qubits _ and _ entanglement purification _
are presented , and their efficiency is investigated in detail . simplicity and flexibility of the idea allow us to apply it to various kinds of settings in quantum information and computation , and would provide us with useful and practical methods of state preparation . |
address the widespread problem of how to take into account differences in standards , confidence and bias in assessment panels , such as those evaluating research quality or grant proposals , employment or promotion applications and classification of university degree courses , in situations where it is not feasible for every assessor to evaluate every object to be assessed . a common approach to assessment of a range of objects by such a panel is to assign to each object the average of the scores awarded by the assessors who evaluate that object .
this approach is represented by the cell labelled `` simple averaging '' ( sa ) in the top left of a matrix of approaches listed in table 1 , but it ignores the likely possibility that different assessors have different levels of stringency , expertise and bias @xcite .
some panels shift the scores for each assessor to make the average of each take a normalised value , but this ignores the possibility that the set of objects assigned to one assessor may be of a genuinely different standard from that assigned to another . for an experimental scientist ,
the issue is obvious : _
calibration_. one is to seek to calibrate the assessors beforehand on a common subset of objects , perhaps disjoint from the set to be evaluated @xcite .
this means that they each evaluate all the objects in the subset and then some rescaling is agreed to bring the assessors into line as far as possible .
this would not work well , however , in a situation where the range of objects is broader than the expertise of a single assessor .
also , regardless of how well the assessors are trained , differences between individuals assessments of objects remain in such ad hoc approaches @xcite .
.panel assessment methods : the matrix of four approaches according to use of calibration and/or confidences .
simple averaging ( sa ) is the base for comparisons .
fisher s iba does not deal with varying degrees of confidence and the confidence - weighted averaging does nt achieve calibration .
the method proposed herein ( cwc ) accommodates both calibration and confidences . [ cols= " < ,
< , < " , ] on this basis , the most precise results are given by cwc .
none of them are very precise , however .
a posterior uncertainty of 8 means that we should consider values for the objects to have a @xmath0 chance of differing by more than 8 from the outputted values .
this means that for iba and cwc , only the top three proposals of table [ ranks ] are reasonably assured of being in the top ten .
as the object of the competition was only to choose the best 10 proposals to fund , rather than assign values to each proposal , it might have been more appropriate to design just a classifier system ( with a tunable parameter to make the right number in the `` fund '' class ) but our goal was to use it as a test of cwc .
the fact that three different methods with roughly equal evidence lead to drastically different allocation of the grants , and with large posterior uncertainties , highlights that better design of the panel assessment was required .
a moral of our analysis is that to achieve a reliable outcome , the assessment procedure needs substantial advance design .
we continue a discussion of design in appendices c and f. we also tested the method on undergraduate examination results for a degree with a flexible options system @xcite and on the assessment of a multi - lecturer postgraduate module . in the former case , as surrogates for the confidences in the marks we took the number of credit accumulation and transfer scheme ( cats ) points for the module , which indicate the amount of time a student is expected to devote to the module ( for readers used to the european credit transfer and accumulation system , 2 cats points are equivalent to 1 ects point ) .
the amount of assessment for a module is proportional to the cats points .
if it can be regarded as consisting of independent assessments of subcomponents , e.g. one per cats point , with roughly equal variances , then the variance of the total score would be proportional to the number of cats points .
as the score is then normalised by the cats points , the variance becomes inversely proportional to the cats points , making confidence directly proportional to cats points .
the outcome indicated significant differences in standards for the assessment of different modules , but as most modules counted for 15 or 18 cats , this was not a strong test of the merits of including confidences in the analysis , so we do not report on it here . for the postgraduate module
, there were four lecturers plus module coordinator , who each assessed oral and written reports for some but not all of the students , according to availability and expertise ( except the coordinator assessed them all ) .
each assessor provided a score and an uncertainty for each assessment .
the results were combined using our method and the resulting value for each student was reported as the final mark .
the lecturers agreed that the outcome was fair .
we have presented and tested a method to calibrate assessors , taking account of differences in confidence that they express in their assessments . from a test on simulated data we found that calibration with confidence ( cwc ) generated closer estimates of the true values than additive ncomplete lock nalysis or simple veraging . a test on some real data , suggesting that the assessment procedure for that context needed more robust design .
nevertheless , cwc came ahead on posterior precision .
there are a number of refinements which one could introduce to the core method .
these include how to deal with different types of bias , different scales for confidence , different ways to remove the degeneracy in the equations , how to deal with the endpoints on a marking scale , and how to choose the assessment graph .
some suggestions are made in the appendices , along with mathematical treatment of the robustness of the method and of computation of the bayesian evidence for the models .
an advantage of type of calibration is that it does not produce the artificial discontinuities across field boundaries that tend to arise if the domain is partitioned into fields and evaluation in each field carried out separately .
we suggest that a method such as this , which takes into account declared confidences in each assessment , is well suited to a multitude of situations in which a number of objects is assessed by a panel .
we are grateful to the mathematics department , university of warwick , for providing us with examination data to perform an early test of the method , to the applied mathematics research centre , coventry university for funding to make a professional implementation of the method and to marcus ong and daniel sprague of spectra analytics for producing it .
we also thank john winn for pointing us to the sigkdd09 method , and david mackay for pointing us to the nips method .
software implementing the method is free to download from the website + .
software and data for the two case studies are available from .
rm conceived and developed the theory .
sp tested it using an early case study .
rl performed case study 1 .
rk performed case study 2 .
rm , sp , rk and rl discussed and interpreted the results and wrote the paper .
the work of rm was supported by the esrc under the network on integrated behavioural science ( es / k002201/1 ) and the centre for evaluation of complexity in the nexus ( es / n012550/1 ) .
rk was supported by the eu marie curie irses network pirses - ga-2013 - 612707 dionicos - dynamics of and in complex systems funded by the european commission within the fp7-people-2013-irses programme ( 2014 - 2018 ) .
we have no competing interests .
xxx meadows m. can we predict who will be a reliable marker ?
manchester : aqa centre for education research and policy , 2006 .
bayesian methods for calibration of examiners .
british journal of mathematical and statistical psychology ( 1981 ) * 34 * , 213 - 223 .
ns t , brockhoff p and tomic o. statistics for sensory and consumer science .
wiley , chicester , 2010 .
fisher ra .
an examination of the different possible solutions of a problem in incomplete blocks .
annals of eugenics ( 1940 ) * 10 * , 5275 .
giesbrecht fg .
analysis of data from incomplete block designs .
biometrics ( 1986 ) * 42 * , 437448 .
flach pa , spiegler s , golenia b , price s , guiver j , harbrich r , graepel t , zaki mj .
novel tools to streamline the conference review process ; experiences from sigkdd09 , + http://research.microsoft.com/pubs/122784/reviewercalibration.pdf guiver , j. calibrating reviews of conference submissions .
+ http://blogs.msdn.com/b/infernet_team_blog/archive2011/09/30/calibrating-reviews-of-conference-submissions.aspx platt j , burges c. regularised least squares to remove reviewer bias .
http:// research.microsoft.com/en-us/um/people/cburges/papers/reviewerbias.pdf ge h , welling m , ghahramani z , a bayesian model for calibrating conference review scores .
http://mlg.eng.cam.ac.uk/hong/nipsrevcal.pdf , 2013 .
thorngate w , dawes rm and foddy m. judging merit .
psychology press , new york , 2008 .
hubbard dw . how to measure anything .
wiley , 2007 , 2010 , 2014 .
matrix analysis and applied linear algebra .
siam philadelphia 2001 .
golub gh and van loan cf . matrix computations .
johns hopkins university press , baltimore , 1996 .
parker s. a test of a method for calibration of assessors .
final year undergraduate project report , university of warwick , april 2014 .
chung frk , spectral graph theory ( am math soc , 1996 ) mackay djc , information theory , inference and learning algorithms ( cambridge univ press , 2003 ) .
song t , wolfe ew , hahn l , less - petersen m , sanders r and vickers d. relationship between rater background and rater performance .
+ http://researchnetwork.pearson.com/wp-content/uploads/ song_raterbackground_04_21_2014.pdf fuchs d and fuchs ls .
test procedure bias : a meta - analysis of examiner familiarity effects .
review of educational research ( 1986 ) * 56 * , 243 - 262 .
we motivated the model by proposing that the noise terms be of the form @xmath1 with the @xmath2 independent zero - mean random variables with unit variance , so that the @xmath3 are standard deviations .
nevertheless , multiplying all the confidences by the same number does not change the results of the least squares fit , nor our quantifications of robustness ( appendices c and d ) .
thus the @xmath2 can be taken to have any variance @xmath4 , as long as it is the same for all assessments .
it is only ratios of confidences that have significance .
the fitting procedure can be extended to infer a best fit value for @xmath4 . even if the assessors provide confidences based on assuming @xmath5 , the best fit for @xmath4 is not 1 in general .
assuming independent gaussian errors , the maximum likelihood value for @xmath4 comes out to be @xmath6 where @xmath7 is the residual from the least squares fit ( @xmath8 for @xmath9 and @xmath10 is the total number of assessments . the posterior distribution for @xmath4 , given a prior distribution , is obtained in appendix d.
we can remove the degeneracy in the equations ( [ eq : system1 ] ) and ( [ eq : system1b ] ) in different manners equation ( [ bias ] ) used here .
indeed , use of ( [ bias ] ) can lead to an average shift from the scores to the true values .
this does not matter if only a ranking is required , but if the actual values are important , then a better choice of degeneracy - breaking condition is needed .
a preferable confidence - weighted degeneracy - breaking condition is @xmath11 which from ( [ eq : system1b ] ) automatically implies @xmath12 , thus avoiding the possibility of such systematic shifts . from a theoretical perspective
, however , the best choice of degeneracy - breaking condition is to choose a reference value @xmath13 ( think of a notional desired mean ) and require @xmath14 where @xmath15 using the notation in ( [ eq : vob ] ) and ( [ eq : voc ] ) this can equivalently be written as @xmath16 to reduce the possible average shift from confidence - weighted average scores to true values , the reference value @xmath13 should be chosen near the confidence - weighted average score @xmath17 choosing @xmath18 exactly equal to @xmath19 gives ( [ eq : avbias ] ) , which makes the confidence - weighted average bias come out to 0 and the confidence - weighted average value come out to @xmath19 .
we will show in appendix c , however , that the results are a factor @xmath20 more robust to changes in the scores if @xmath13 is chosen to be fixed rather than dependent on the scores .
here we present our approach to the quantification of the robustness of our method to small changes in the scores , using norms that take into account the confidences . for @xmath21 , define the operator @xmath22 by @xmath23 , \ ] ] as a shorthand for the definitions in equations ( [ referee1a ] ) and ( [ referee1b ] ) , so that the equations ( [ eq : system1 ] , [ eq : system1b ] ) can be written as @xmath24 = k s .\ ] ]
thus , if a change @xmath25 is made to the scores , we obtain changes @xmath26 , @xmath27 of magnitude bounded by @xmath28 where @xmath29 is defined by restricting the domain of @xmath30 to ( [ bias ] ) and its range to @xmath31 , and appropriate norms are chosen . in this appendix
, we propose that appropriate choices of norms are @xmath32 @xmath33 and the associated operator norm from scores to results for @xmath34 . with the confidence - weighted degeneracy - breaking condition @xmath35 ( [ eq : avbias ] ) instead of ( [ bias ] ) we obtain @xmath36 where @xmath37 is the second smallest eigenvalue of a certain matrix @xmath38 formed from the confidences ( see ( [ eq : m ] ) ) .
in particular , this gives @xmath39 the factor of @xmath20 can be removed if one switches to an ideal degeneracy - breaking condition as in ( [ tightest ] ) of appendix b. as a consequence , to maximise the robustness of the results , the task for the designer of @xmath40 is to make none of the @xmath41 much smaller than the others and to make @xmath37 significantly larger than 0 .
the former is evident ( no object should receive significantly less assessment or less expert assessment than the others ) .
the latter is the mathematical expression of how well connected is the graph @xmath42 ( equivalently @xmath43 ) . to design the graph @xmath42 requires a guess of the confidence levels that assessors are likely to give to their assessments ( based on knowing their areas of expertise and their thoroughness or otherwise ) and a compromise between assigning an object to only the most expert assessors for that object and the need to achieve a chain of comparisons between any pair of assessors .
we now go into detail , derive the above bounds and describe some computational shortcuts . one can measure the size of a change @xmath44 to a score @xmath45 by comparing it to the declared uncertainty @xmath3 .
thus we take the size of @xmath44 to be @xmath46 .
we propose to measure the size of an array @xmath25 of changes @xmath44 to the scores by the square root of the sum of squares of the sizes of the changes to each score , as in ( [ eq : scoresnorm ] ) .
supremum or sum - norms could also be considered but we will stick to this choice here .
it is also reasonable to measure the size of a change @xmath47 to a true value @xmath48 by comparing it to the uncertainty implied by the sum of confidences in the scores for object @xmath49 .
thus the size of @xmath47 is defined to be @xmath50 , where @xmath41 is the total confidence in the assessment of object @xmath49 .
similarly , we measure the size of a change @xmath51 in bias @xmath52 by @xmath53 where @xmath54 is the total confidence expressed by a given assessor .
finally , we measure the size of a change @xmath55 to the vector of values and biases by the square root of sum of squares of the individual sizes , as in ( [ eq : resultsnorm ] ) .
the size of the operator @xmath56 is measured by the operator norm from scores to results , i.e. @xmath57 the operator @xmath56 is equivalent to orthogonal projection with respect to the norm ( [ eq : scoresnorm ] ) from the scores to the subspace @xmath58 of the form @xmath59 with a degeneracy - breaking condition to eliminate the ambiguity in direction of the vector @xmath60 . the tightest bounds in ( [ bound ] )
are obtained by choosing the degeneracy - breaking condition to correspond to a plane perpendicular to this vector with respect to the inner product corresponding to equation ( [ eq : resultsnorm ] ) .
thus we choose degeneracy - breaking condition ( [ tightest ] ) .
1ex * theorem * : for a connected graph @xmath42 and with the degeneracy - breaking condition ( [ tightest ] ) , the size of the change @xmath55 resulting from a given array of changes @xmath25 in scores is bounded by @xmath61 where @xmath37 is the second smallest eigenvalue of the matrix @xmath62 , \label{eq : m}\ ] ] @xmath63 @xmath64 , @xmath65 are the numbers of assessors and objects respectively , and for @xmath66 @xmath67 is the identity matrix of rank @xmath68 .
1ex * proof * : firstly , the orthogonal projection in metric ( [ eq : scoresnorm ] ) from @xmath69 to the subspace @xmath58 never increases length . secondly , if @xmath70 with @xmath71 then @xmath72 where @xmath73 is the vector with components @xmath74 then , because we restricted to the orthogonal subspace to the null vector in results - norm and @xmath38 is non - negative and symmetric , @xmath75 where index @xmath76 ranges over all objects and assessors .
positivity of @xmath37 holds as soon as the graph @xmath42 is connected , because @xmath38 is a transformation of the weighted graph - laplacian to scaled variables @xcite , so dividing by @xmath37 and taking the square root yields the result .
@xmath77 1ex the computation of the eigenvalue @xmath37 of @xmath38 can be reduced from dimension @xmath78 to dimension @xmath64 by 1ex * proposition * : if @xmath79 , the second smallest eigenvalue @xmath37 of @xmath38 is related to the second largest eigenvalue @xmath80 of @xmath81 by @xmath82 if @xmath83 and @xmath84 then @xmath85 . if both are 1 then @xmath86 .
1ex * proof * : the equations for an eigenvalue - eigenvector pair @xmath87 of @xmath38 are @xmath88 applying @xmath89 to the first equation , multiplying the second by @xmath90 , and then substituting for @xmath91 in the second yields @xmath92 thus either @xmath93 or @xmath94 is an eigenvalue @xmath95 of @xmath81 .
in the first case , equation ( [ eq : evec ] ) implies @xmath96 , so if @xmath97 then @xmath94 is an eigenvalue of @xmath98 .
conversely , if @xmath99 is an eigenvalue - eigenvector pair for @xmath98 with @xmath100 then @xmath101 because @xmath81 is non - negative , so put @xmath102 to see that @xmath103 is an eigenvector of @xmath38 with eigenvalue @xmath104 .
if @xmath105 and @xmath106 then @xmath107 is an eigenvalue of @xmath38 with eigenvector @xmath103 for any @xmath108 with @xmath109 , e.g. @xmath110 . thus there is a two - to - one correspondence between eigenvalues @xmath111 of @xmath38 not equal to 1 and positive eigenvalues @xmath95 of @xmath81 ( counting multiplicity ) : @xmath104 .
any remaining eigenvalues are 1 for @xmath38 and 0 for @xmath98 .
the degeneracy gives an eigenvector @xmath112 of @xmath38 with eigenvalue 0 and it corresponds to an eigenvalue 1 of @xmath98 .
all other eigenvalues of @xmath38 are non - negative because @xmath38 is .
all other eigenvalues of @xmath98 are less than or equal to 1 by the cauchy - schwarz inequality .
so if the second largest eigenvalue @xmath80 of @xmath98 ( counting multiplicity ) is positive then the second smallest eigenvalue @xmath37 of @xmath38 ( counting multiplicity ) is @xmath113 .
if @xmath114 then @xmath115 because existence of @xmath80 implies @xmath116 so @xmath38 has dimension at least 3 and we have only two simple eigenvalues @xmath117 and @xmath118 from the simple eigenvalue 1 of @xmath98 , so @xmath38 must have another one but any other value than 1 would give a positive @xmath80 ; so the same formula holds . if there is no second eigenvalue of @xmath98 (
because @xmath119 ) then if @xmath120 the second largest eigenvalue of @xmath38 must be 1 by the same argument . if both @xmath64 and @xmath65 are 1 then the second largest eigenvalue of @xmath38 is the other one associated with the eigenvalue 1 of @xmath98 , namely 2 .
@xmath77 1ex note that @xmath121 is a similarity transformation of ( [ eq : aweights ] ) . as examples of second eigenvalues ,
putting unit confidences on the graphs in the left column of figure [ fig1:graphs ] we calculate @xmath122 for cases ( a),(b),(c ) in the right column , giving @xmath123 , respectively .
finally , a user may prefer to use the degeneracy - breaking condition ( [ eq : avbias ] ) rather than ( [ tightest ] ) , perhaps out of uncertainty about what value of @xmath18 to use . or a user may be happy to use ( [ tightest ] ) with @xmath18 equal to the confidence - weighted average score , but wants @xmath18 to follow this average score if changes are made to the scores .
that comes out equivalent to using ( [ eq : avbias ] ) .
so we extend our discussion of robustness to treat this case .
we find it makes the bounds increase by a factor of only @xmath20 .
1ex * proposition * : for @xmath42 connected and using degeneracy - breaking condition ( [ eq : avbias ] ) , the size of @xmath55 resulting from changes @xmath25 to the scores is at most @xmath124 .
1ex * proof * : if the degeneracy - breaking condition ( [ tightest ] ) gives a change @xmath55 for a change @xmath25 to the scores , then switching to degeneracy - breaking condition ( [ eq : avbias ] ) just adds an amount @xmath68 of the null vector @xmath125 to achieve @xmath126 , i.e. @xmath127 in the results metric , the null vector has length @xmath128 .
thus the correction has length @xmath129 . using the condition ( [ tightest ] )
we can write @xmath130 which one can recognise as one half of the inner product of @xmath131 with @xmath55 in results - norm , so it is bounded by @xmath132 .
thus the length of the correction vector is at most that of @xmath55 .
the correction is perpendicular to @xmath55 , thus the vector sum has length at most @xmath133 .
@xmath77 1ex one may also ask about robustness with respect to changes in the confidences @xmath134 . if an assessor declares extra high confidence for an evaluation , for example , that can significantly skew the resulting @xmath135 and @xmath136 . the analysis is more subtle , however , because of how the @xmath134 appear in the equations and we do not treat it here .
another point of view on robustness is the bayesian one . from a prior probability on @xmath9 and a model for the @xmath2 ,
one can infer a posterior probability for @xmath9 , whose inverse width tells one how robust is the inference . in the case of
flat prior on @xmath9 , prescribed @xmath4 , gaussian noise , and an affine degeneracy - breaking condition , the posterior is gaussian with mean at the value solving equations ( [ eq : system1 ] ) , ( [ eq : system1b ] ) and the degeneracy - breaking condition , and with covariance matrix related to @xmath29 .
specifically , the posterior probability density for @xmath9 is proportional to @xmath137 constrained to the degeneracy - breaking hyperplane , where @xmath138 using ( [ eq : gtmg ] ) and ( [ eq : r ] ) , this can be written as @xmath139 with @xmath55 being the deviations of @xmath9 from the least squares fit .
thus the covariance matrix in these scaled variables is @xmath140 . using the degeneracy - breaking condition ( [ tightest ] ) or equivalently ( [ eq : tightest ] )
, we obtain widths @xmath141 for the posterior on @xmath73 in the eigendirections of @xmath38 , where @xmath142 are the positive eigenvalues of @xmath38 .
thus the robustness of the inference is again determined by @xmath37 , but scaled by @xmath143 .
a slightly more sophisticated approach is to consider @xmath4 to be unknown also . given a prior density @xmath144 for @xmath4 ( which could be peaked around 1 if the assessors are assigning confidences via uncertainties , but following jeffreys would be better chosen to be @xmath145 if there is no information about the scale for the confidences ) , the posterior density for @xmath146 is proportional to @xmath147 where again @xmath10 is the number of assessments .
the maximum of the posterior probability density is determined by the least squares fit for @xmath9 ( which is independent of @xmath4 ) and the following equation for @xmath4 : @xmath148 for @xmath10 large , the peak of the posterior has @xmath4 near the previously determined maximum likelihood value @xmath149 . , taking jeffreys prior , the peak is at @xmath150 . integrating over @xmath4 ( with jeffreys prior )
e find the marginal posterior for @xmath9 to be proportional to @xmath151 incorporating an affine degeneracy - breaking condition , this is a @xmath152-variate student distribution with @xmath153 degrees of freedom .
its covariance matrix is @xmath154 with @xmath155 and @xmath156 interpreted by imposing the chosen degeneracy - breaking condition .
so for the degeneracy - breaking condition ( [ tightest ] ) , the robustness of the inference is given by widths @xmath157 for @xmath158 , in the eigendirections of @xmath38 on @xmath73 .
in particular , the confidence - weighted root mean square uncertainty @xmath159 for the components of the vector @xmath9 is @xmath160 where @xmath161 denotes the trace and , again , @xmath156 is interpreted by restricting to the degeneracy - breaking plane .
marginal posteriors for each @xmath48 and @xmath52 can be extracted , but it must be understood that in general they are significantly correlated . for the case of simple averaging , the root mean - square posterior uncertainty in the values , weighted by the numbers @xmath162 of assessors for object @xmath49 , is @xmath163 where @xmath164 is defined in ( [ eq : rsa ] ) of appendix e. this can be derived in an analogous fashion to ( [ eq : sigma ] ) via a student distribution again , but with @xmath165 .
here we describe the method used in case study 2 to compare the three models .
bayesian model comparison is based on computing how much evidence there is for each proposed model , e.g. ch.28 of @xcite .
the evidence for a model @xmath38 given data @xmath166 is @xmath167 .
given strength of belief @xmath168 in model @xmath38 prior to the data ( relative to other models ) , one can multiply it by the evidence to obtain the posterior strength of belief in model @xmath38 .
it is convenient to replace multiplication by addition , thus we define the log - evidence @xmath169 if the model @xmath38 has free parameters @xmath111 then @xmath170 where @xmath171 is a prior probability density on @xmath111 .
let there be @xmath65 objects , @xmath64 assessors , let @xmath45 be the score returned by assessor @xmath172 for object @xmath49 , @xmath134 the confidence in this score in the case of calibration with confidence , @xmath69 be the collection of scores and @xmath10 be their number .
first we compute the evidence for simple averaging ( sa ) .
then we treat calibrate with confidence ( cwc ) and lastly incomplete block analysis ( iba ) because it is a special case of cwc . for simple averaging ( sa ) ,
the model is that @xmath173 for some unknown vector @xmath135 of `` true '' values @xmath48 , with @xmath174 iid normal @xmath175 for some unknown variance @xmath4 .
then the probability density for the scores @xmath69 is @xmath176 with the product and sum being over the assessments that were carried out . to work out the evidence for sa
the model must include a prior probability density for @xmath135 and @xmath4 .
the simplest proposal would be @xmath177 on @xmath178 $ ] , @xmath179 $ ] , where @xmath180 and @xmath181 .
this is the product of a `` box '' prior on @xmath135 and jeffreys prior on @xmath4 ( truncated to an interval and normalised ) . for comparison with the other models ,
however , it is easier to replace the box prior on @xmath135 by a `` ball '' prior , giving @xmath182 on @xmath183 for some anticipated average score @xmath184 and upper estimate of the width @xmath185 of the distribution of values @xmath48 .
the normalisation is @xmath186 where @xmath42 is the gamma function . for @xmath187
is it reasonable to choose @xmath188 where @xmath189 is the smallest change any assessor could contemplate .
for @xmath190 it is reasonable to choose @xmath191 . for each object @xmath49 , @xmath192 where @xmath162 is the number of assessors for object @xmath49 , @xmath193 is the mean of their scores , and the residual @xmath194 thus @xmath195 where @xmath196 to integrate this over @xmath135 and @xmath4 , we assume the bulk of the probability distribution lies in the product of the ball and the interval , and so approximate by extending the range of integration to @xmath197 . integrating the exponential over @xmath48 produces a factor @xmath198 thus , integrating over all components of @xmath135 yields @xmath199 integrating this over @xmath4 , we obtain the evidence @xmath200 and the log - evidence @xmath201 for calibrate with confidence ( cwc ) , the model is @xmath202 for some unknown vectors @xmath135 of true values @xmath48 , and @xmath136 of assessor biases @xmath52 , with @xmath2 iid normal @xmath175 for some unknown variance @xmath4 .
the uncertainties @xmath3 correspond to confidences @xmath134 by @xmath203 , which are considered as given ( one could propose a generative model for them too , but that would require further analysis ) . then the probability density for @xmath69 is @xmath204 for prior probability density over the parameters @xmath205 , we want to build in a degeneracy - breaking condition .
we used @xmath206 in our calculation , thus we take prior `` density '' @xmath207 on the product of the balls ( [ eq : ballo ] ) and @xmath208 and interval @xmath209 $ ] , where @xmath210 is the delta function .
here , @xmath211 is an estimated upper bound for the standard deviation of the biases , and the normalisation is @xmath212 note that the interpretation of @xmath4 is not the same as for sa , so one might choose a different prior for it .
for example , if the @xmath3 are fairly accurate values for the uncertainties in the scores then the prior for @xmath4 should be peaked around @xmath5 , but if they are on an undetermined scale a truncated jeffreys prior is sensible .
the only thing is that one might want to choose a different interval for it , but for application to iba where the @xmath213 or to cwc if the @xmath134 are on a scale centred around 1 , such as we have used to translate the quantitative high / medium / low confidence ratings , the same interval should be reasonable .
similarly , one might want to use a different value for @xmath185 if one believes that the spread in values is more due to variation in assessor bias than true value , but in our case we think it reasonable to use the same @xmath185 .
thus @xmath214 again we assume the bulk of this lies in the product of balls and interval , so we approximate its integral by extending the domains to infinity . now
@xmath215 where @xmath216 is the vector with @xmath217 components , @xmath218 , @xmath219 , @xmath220 is any least squares fit to this model ( without loss of generality satisfying the degeneracy - breaking condition ) , the residual @xmath164 is now @xmath221 ( as in ( [ eq : r ] ) ) and @xmath222 is the matrix with block form @xmath223.\ ] ] choose one assessor , say @xmath224 , and integrate over @xmath225 .
this yields @xmath226 with @xmath227 being the remaining components of @xmath216 and @xmath228,\ ] ] where @xmath229 and @xmath230 , restricted to @xmath231 , which takes into account that @xmath232 .
thus the integral over @xmath227 is @xmath233 where @xmath234 .
finally , we integrate over @xmath4 to obtain @xmath235 and the log - evidence is @xmath236 the model for incomplete block analysis ( iba ) is the same as for cwc but taking the confidences @xmath213 for all the assessments .
thus the log - evidence for iba given the scores @xmath69 is @xmath237 with the appropriate changes to @xmath164 and @xmath238 .
one could develop refinements to the basic model ( [ eq : model1 ] ) .
for example , assessors might have not only an additive bias but also different scales , so for example @xmath239 fitting @xmath240 is more complicated , however , than just @xmath241 .
an assessor may have a bias correlated with their confidence @xcite or with some other feature like familiarity @xcite .
assessors may like to give round - number scores . may have different scales for confidence , so their confidences may need calibrating as well as their scores .
another problem is that often assessors are asked to assign scores in a fixed range @xmath242 $ ] , e.g. @xmath243 .
then any model for bias really ought to be nonlinear to respect the endpoints .
one way to treat this is to apply a nonlinear transformation to map a slightly larger interval @xmath244 onto @xmath245 , e.g. @xmath246 or @xmath247 apply our method to the transformed scores , scaling the confidences by the inverse square of the derivative of the transformation , and then apply the inverse transformation to the true " values . on the other hand , it may be inadvisable to specify a fixed range because it requires an assessor to have knowledge of the range of the objects before starting scoring
. thus one could propose asking assessors to use any real numbers and then use equation ( [ eq : model2 ] ) to extract true values @xmath135 .
a simpler strategy that might work nearly as well is to allow assessors to use any positive numbers but then to take logarithms and fit equation ( [ eq : model1 ] ) to the log - scores .
the assessor biases would then be like logarithms of exchange rates .
the confidences would need translating appropriately too .
one issue with our method is that the effect of an assessor who assesses only one object is only to determine their own bias , apart from an overall shift along the null vector @xmath248 for the rest . to rectify
this one could incorporate a prior probability distribution for the biases ( indeed , this was done by @xcite in the form of a regulariser ) .
an interesting future project is to design the graph @xmath42 optimally , given advance guesses of confidences and constraints or costs for the number of assessments per assessor .
`` optimality '' would mean to achieve maximum precision or robustness of the resulting values .
for instance , in each case of figure [ fig1:graphs ] , each assessor has the same amount of work and each object receives the same amount of attention , but ( a ) achieves full connectivity with a resulting value for @xmath37 of @xmath249 , whereas ( b ) achieves moderate connectivity and a smaller value of @xmath250 , and ( c ) is not even connected and has @xmath251 . | frequently , a set of objects has to be evaluated by a panel of assessors , but not every object is assessed by every assessor .
a problem facing such panels is how to take into account different standards amongst panel members and varying levels of confidence in their scores . here
, a mathematically - based algorithm is developed to calibrate the scores of such assessors , addressing both of these issues .
the algorithm is based on the connectivity of the graph of assessors and objects evaluated , incorporating declared confidences as weights on its edges .
if the graph is sufficiently well connected , relative standards can be inferred by comparing how assessors rate objects they assess in common , weighted by the levels of confidence of each assessment . by removing these biases , true " values are inferred for all the objects .
reliability estimates for the resulting values are obtained .
the algorithm is tested in two case studies , one by computer simulation and another based on realistic evaluation data .
the process is compared to the simple averaging procedure in widespread use , and to fisher s additive incomplete block analysis .
it is anticipated that the algorithm will prove useful in a wide variety of situations such as evaluation of the quality of research submitted to national assessment exercises ; appraisal of grant proposals submitted to funding panels ; ranking of job applicants ; and judgement of performances on degree courses wherein candidates can choose from lists of options .
keywords : calibration , evaluation , assessment , confidence , uncertainty , model comparison . |
consider a mesh having one special @xmath1-vertex face and a `` strip '' of faces attached to the special face along the perimeter , such that any vertex of a special face is a vertex of exactly three faces in the strip .
this mesh is called a _
kokotsakis mesh_. kokotsakis meshes with polygonal faces are known in the literature as _ polyhedral surfaces of kokotsakis type_. in this paper we study conditions of flexibility and infinitesimal flexibility for kokotsakis meshes . the first steps in studying such meshes were made in 1932 by a. kokotsakis in his paper @xcite , where he found conditions of infinitesimal flexibility .
recently kokotsakis meshes for @xmath2-gons ( such meshes are also called _ @xmath3-meshes _ ) were used in the study of discrete conjugate nets flexibility .
a. i. bobenko , t. hoffmann , w. k. schief proved in @xcite that a nondegenerate discrete conjugate net is isometrically deformable if and only if all its @xmath4 subcomplexes are isometrically deformable @xmath5-meshes . in
@xcite they also showed that the conjugate nets admit infinitesimal second - order deformations if and only if their reciprocal - parallel surfaces constitute discrete bianchi surfaces ( see also @xcite ) . in this paper we give a geometrical interpretation of infinitesimal flexibility in terms of ceva and menelaus configurations .
further we work with equations that define the set of flexible kokotsakis meshes , in particular we show semi - algebraicity of this set .
the technique developed in this paper allows us to calculate linear independence of the gradients of the first 6 flexibility conditions for a chosen flexible quadrangular kokotsakis mesh .
this means that we can check if the codimension of a stratum of meshes is smaller than or equal to 6 .
the computational complexity of the problem does not allow us to proceed further .
we remind the reader that the codimension of the set of flexible voss meshes as well as that of symmetric flexible meshes is 8 , for information on voss surfaces we refer to @xcite .
so it is most probable that the codimensions of irreducible components coincide , and equal 6 , 7 , or 8 .
if we are lucky and these codimensions are exactly 8 , than the geometric solution of the problem can be obtained by enumeration of all possible components . to check the conditions at a single point we introduce more simple `` flexibility vector calculus '' conditions . in the case of @xmath5-meshes
we now able to calculate the 8 first such conditions .
finally , for the case of quadrangular kokotsakis @xmath3-meshes we show that the property of the mesh to be flexible depends only on the shapes of faces but not on the geometry of the mesh in @xmath6 .
we propose an algorithm to write these formulae explicitly .
* organization of the paper .
* we start in section 1 with showing an infinitesimal flexibility condition in algebraic and geometric terms .
section 2 studies properties of the set of flexible meshes .
we describe equations defining this set . in section
3 we introduce a technique to write flexibility conditions in terms of angles of the faces of meshes .
we conclude the paper in section 4 with some open questions for further study . *
general notation . * to avoid the annoying description of multiple cases we consider the index @xmath7 belonging to the cyclic group @xmath8 .
we also choose a partial _ ordering _
@xmath9 for this group .
we denote the vertices of the special face in the middle of the mesh by @xmath10 , @xmath11 , preserving their order at the boundary .
let @xmath12 . notice that by definition we have @xmath13 .
consider a vertex @xmath10 ( @xmath14 ) .
we denote vertices @xmath15 and @xmath16 such that the four faces meeting at @xmath10 contain angles @xmath17 , @xmath18 , @xmath19 , and @xmath20 .
we denote these angles by @xmath21 , @xmath22 , @xmath23 , and @xmath24 , respectively .
let @xmath25 and @xmath26 .
@xmath27 _ remark .
_ without loss of generality we assume that the special face @xmath28 is fixed , while the other @xmath29 faces may move
. denote by @xmath30 and by @xmath31 $ ] the scalar and vector products , respectively .
in this subsection we recall some results on infinitesimal flexibility from the work @xcite by a. kokotsakis .
we rewrite his results in a slightly modified way more convenient for us .
consider a @xmath32-mesh with a common vertex @xmath33 .
let us fix a face @xmath34 .
let _ bav _ , _ vaw _ , and _ wac _ be the other three faces .
denote by @xmath35 , @xmath36 , @xmath37 , and @xmath38 the vectors @xmath39 , @xmath40 , @xmath41 , and @xmath42 respectively .
this mesh is flexible in general position and has one degree of freedom . [
basis ] suppose that the vectors @xmath37 and @xmath38 are not collinear and the vectors @xmath35 and @xmath36 are not in the plane spanned by the vectors @xmath37 and @xmath38 .
then for any infinitesimal motion @xmath43 of the described @xmath32-mesh there exists a non - zero real @xmath44 such that @xmath45}{(b , v , w)}\\ \dot{w}&=&\displaystyle \lambda\frac{[c , w]}{(c , v , w)}\\ \end{array } \right . .\ ] ] any infinitesimal motion is a solution of the following linear system with 6 variables ( coordinates of @xmath46 and @xmath47 ) and 5 equations : @xmath48 on one hand the solution mentioned in the statement of the proposition satisfies this system . on the other hand , if the conditions on vectors @xmath35 , @xmath36 , @xmath37 , and @xmath38 hold then the rank of the linear system is 1 ( this follows from direct calculations in coordinates ) . from proposition [ basis ]
we immediately have the following statement ( this statement is another formulation of the result in @xcite by a. kokotsakis ) . for a kokotsakis mesh @xmath49 put by definition @xmath50 [ flex ]
a kokotsakis mesh @xmath49 is infinitesimally flexible if and only if @xmath51 .
consider points @xmath52 in a plane .
we denote this plane by @xmath53 .
let @xmath54 denote the intersection of the plane containing the points @xmath10 , @xmath15 , and @xmath16 with the plane @xmath53 .
we suppose that the lines @xmath54 and @xmath55 are not parallel , and denote their intersection point by @xmath56 . for a @xmath5-mesh constructed with planar faces the infinitesimal flexibility conditions are shown in the following proposition . for more information see the work @xcite by h. pottmann , j. wallner and the book @xcite by r. sauer .
consider a mesh in general position .
then it is infinitesimally flexible if and only if the lines @xmath57 , @xmath58 , and @xmath59 intersect in one point @xmath60i.e .
the points @xmath10 , @xmath56 , @xmath61 are aligned in a desargues configuration@xmath62 .
let us generalize this statement for an arbitrary kokotsakis mesh with planar polygons as faces .
for @xmath11 we define real numbers @xmath63 from the following affine equations : @xmath64 [ relations ] the closed broken line @xmath65 satisfies the generalized ceva - menelaus condition , i.e. @xmath66 if there is one point @xmath56 `` at infinity '' one should replace @xmath67 in the product . here
@xmath63 and @xmath68 denotes the absolute values of the vectors @xmath69 and @xmath70 respectively .
we take @xmath71 if the directions of these two vectors are opposite and @xmath72 if they coincide .
we leave for the reader the other cases of some of the vertices @xmath56 being `` at infinity '' as an exercise .
choose some orientation on the plane @xmath53 .
for an ordered couple @xmath73 of vectors denote @xmath74 denote by @xmath75 the ( non - oriented ) distance from the point @xmath76 to the line @xmath77 .
then for any @xmath78 we have : @xmath79 therefore , @xmath80 the second equality holds by the law of sines for the triangles @xmath81 for @xmath11 .
denote by @xmath82 $ ] the point of intersection of the lines passing through @xmath83 and @xmath84 respectively .
denote @xmath85 .
let @xmath86 $ ] , and @xmath87 $ ] , for @xmath88 .
[ geom]*geometric condition i. * the condition @xmath89 is equivalent to the condition that the points @xmath90 and @xmath91 coincide @xmath60see figure [ pentagon].i@xmath62 .
@xmath92 in the case @xmath93 we have : _ the lines @xmath94 , @xmath95 , and @xmath96 intersect at one point @xmath60ceva configuration@xmath62_. in the case @xmath97 we have : _ the lines @xmath57 , @xmath58 , and @xmath59 intersect at one point @xmath60desargues configuration@xmath62_. we prove the proposition by induction on @xmath1 . by ceva s theorem and theorem [ relations ]
the statement holds for all configurations with @xmath93 . for pairwise
distinct collinear points @xmath33 , @xmath98 , @xmath99 denote @xmath100$}\\ -1 & \hbox{otherwise } \end{array } \right . .\ ] ] suppose the statement holds for all configurations with @xmath97 . by ceva s theorem in the triangle
@xmath101 we have : @xmath102 then by induction the statement is true for the broken line @xmath103 .
[ geom2]*geometric condition ii . *
the condition @xmath104 is equivalent to the condition that the points @xmath105 , @xmath106 , and @xmath90 are collinear @xmath60see figure [ pentagon].ii@xmath62 . in the case
@xmath93 we have : _ the points @xmath107 , @xmath108 , and @xmath109 are contained in one line _
@xmath60menelaus configuration@xmath62 .
we prove the proposition by induction on @xmath1 . by menelaus theorem and theorem [ relations ] the statement holds for all configurations with @xmath93 .
the induction step is the same as in the proof of proposition [ geom2 ] , so we omit it here .
the condition of corollary [ flex ] can be rewritten as geometric condition i in the case of even @xmath1 , and as geometric condition ii in the case of odd @xmath1 .
any kokotsakis mesh with a triangle in the middle , in general position , is not infinitesimal flexible @xmath60and therefore it is not flexible@xmath62 .
in particular we have the following conclusion .
consider a closed mesh in general position ( i.e. there are no 3 vertices lying in one line ) whose edge - graph is 4-valent . then the mesh is not flexible if at least one of its faces is a triangle .
actually this means that the euler characteristic of the surface is non - positive .
so a flexible mesh in general position is not homeomorphic to the sphere or to the projective plane .
first we define a configuration space of all kokotsakis meshes whose central face is an @xmath1-gon .
any kokotsakis mesh contains @xmath110 points .
so the configuration space of all possible vertex configurations is @xmath111 .
we will not consider configurations for which some face has three consequent vertices in a line , or which has faces with coinciding vertices .
denote this set by @xmath112 .
finally , the _ configuration space of all kokotsakis meshes _ is by definition @xmath113 , denote it by @xmath114 .
notice that we are not restricted to the case of planar faces in this section .
consider the vector field : @xmath115\quad \hbox{and } \\
\dot{w_i}&=&\displaystyle \left(\prod\limits_{j=1}^{i } \frac{(a_{j-1},v_j , w_j)}{(a_j , v_j , w_j)}\right)[a_i , w_i]\quad \hbox{for } i=1,\ldots , n.\\ \end{array}\ ] ] and denote it by @xmath116 .
this vector field is clearly everywhere defined in @xmath114 .
further when , in this section , we speak of a motion we assume that the special face in the middle is fixed while the others are moving . consider a flexible mesh in @xmath114 . the motion trajectory of the mesh in some small neighborhood is contained in some integral curve of the vector field @xmath116 . by proposition [ basis ]
a tangent vector for the trajectory of a flexible mesh @xmath49 in @xmath114 is proportional to the vector @xmath117 .
this implies the statement of the theorem .
denote the set of all flexible meshes by @xmath118 .
let us consider an infinite sequence @xmath119 of algebraic functions on @xmath111 , defined inductively : + _ i@xmath62 .
_ take @xmath120 to be the numerator of the rational function @xmath121 . +
_ ii@xmath62 . _
take @xmath122 to be the numerator of the rational function @xmath123 .
we ignore denominators since they are nonzero algebraic functions on @xmath114 .
the closure of @xmath118 in @xmath114 is a semialgebraic set coinciding with @xmath124 notice that the vector field @xmath117 is non - zero at all the points of @xmath114 . by proposition [ basis ] ,
for a flexible mesh @xmath49 not contained in @xmath112 the velocity vector is proportional to the vector @xmath117 , and therefore satisfies all the conditions @xmath125 . on the other hand all functions @xmath122 are algebraic and therefore their common locus is an algebraic variety ( by hilbert s basis theorem it is sufficient to take only a finite number of such equations ) .
therefore in a nonsingular point of this variety we have all the existence conditions for a finite motion along the field @xmath117 ( see in @xcite ) .
since the set of singularities of an algebraic variety is nowhere dense in this variety , the closure of the set of nonsingular points of a variety is the variety itself . in practice
it is quite hard to calculate the above @xmath122 explicitly . to check some given configuration
we recommend a slightly different technique . since @xmath126 along any trajectory
, we immediately have many conditions of flexibility : @xmath127 , where the derivatives are taken with respect of the time parameter of the vector field .
we recommend to simplify the obtained function after each round of differentiation using previous ones .
let us introduce basic ingredients to calculate derivatives .
we use a prime to indicate differentiation .
[ cols= " < " , ]
in this section we describe a technique to verify the flexibility of a ( @xmath128)-mesh with planar faces in terms of inner geometry of a mesh , i.e. , in terms of the fixed angles between the edges of the same faces .
consider a @xmath5-mesh with a planar quadrangle @xmath129 in the middle .
denote the angles between the faces adjacent to @xmath130 by @xmath131 ( @xmath14 ) .
the following relation holds for any vertex @xmath10 @xmath132 denote this equation by @xmath133 .
note that the values of the angles do not depend on the lengths of vectors @xmath134 , @xmath135 , @xmath136 , @xmath137 .
so we can choose them equal @xmath138 .
let us make calculations in the coordinate system where @xmath139 , and @xmath140 .
then we have @xmath141 the formula of the lemma coincides with @xmath142 and therefore holds .
we get the desired equations in 4 steps .
_ step 1 . _
denote @xmath143 by @xmath63 .
the first step is to get rid of @xmath144 in formula @xmath133 .
this can be done by putting the corresponding term to the right and the rest to the left , than squaring both sides , and finally substituting @xmath145 .
denote the resulting equation by @xmath146 .
it leads to @xmath147 where @xmath148 _ step 2 .
_ now we recursively evaluate @xmath149 as a linear function in @xmath63 whose coefficients are rational functions of @xmath150 .
we will need only the formulae for @xmath151 and @xmath152 .
_ we use @xmath153 and the formula for @xmath151 given in step 2 to evaluate @xmath154 as a function of @xmath155 and @xmath156 , and analogously we employ @xmath157 with the formula for @xmath152 given in step 2 to evaluate @xmath158 as a function of @xmath155 and @xmath156 .
we get @xmath159 _ step 4 .
_ now for @xmath155 and @xmath156 we have the system @xmath160 first , note that the functions @xmath161 and @xmath162 are rational in variables @xmath155 and @xmath156 .
so we can restrict ourselves to consider the numerators .
denote these numerators by @xmath163 , and @xmath164 .
the functions @xmath120 and @xmath165 are polynomials of degree 10 , and we need to check if these two polynomials define a non - discrete set .
this happens only if @xmath120 and @xmath165 have a common nonconstant factor .
take a resultant of @xmath120 and @xmath165 considered as polynomials in @xmath156 .
this resultant denote it by @xmath166 is a polynomial in @xmath155 .
a mesh in @xmath114 is flexible if and only if all the coefficients of the polynomial @xmath166 equal zero .
we give an outline of the proof .
it is clear that if a mesh is flexible then the set of common solutions of @xmath167 for @xmath61 is nondiscrete and therefore one of the equations can be eliminated . by proposition [ basis ] in the case of meshes in @xmath114
all the @xmath63 s are nonconstant while the mesh is moving , so the resultant @xmath166 is identically zero .
suppose now that all the coefficients of the polynomial @xmath166 equal zero , then there are at most three independent equations among @xmath167 for @xmath61 .
the common set of solutions of the equations @xmath167 is almost always at least one - dimensional .
the dimension of this set is zero only if the corresponding mesh is a limit of a sequence of flexible meshes with possible motion curves shrinking to this mesh .
such mesh is infinitesimally flexible with more than one degrees of freedom . by proposition [ basis ]
such meshes are not in @xmath114 .
flexibility of meshes in @xmath114 is determined by the shapes of faces @xmath60but not of an embedding in the space as a mesh@xmath62 . to simplify symbolic calculations we recommend to compute the resultants of @xmath168 and @xmath169 for @xmath170 .
if all 101 ( = @xmath171 ) resultants evaluate to zero , than the polynomial @xmath166 , which has at most 100 zeroes , has all zero coefficients . finally , we give an algorithm that calculates the functions @xmath172 and @xmath173 ( as @xmath174 $ ] and @xmath175 $ ] in the output below ) .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .... twoequations:=proc(alpha , beta , gama , phi , n ) local i , c , r , t , f , tn , td ; for i from 1 to 4 do c[1,i]:=(cos(alpha[i])*cos(beta[i])*cos(phi[i])-cos(gama[i]))^2- sin(beta[i])^2*sin(phi[i])^2 ; ... # here we write all the formulae for c[2,i], ... ,c[9,i ] from step 1 end do : tn[2]:=-((c[6,2]+c[8,2]*t[1]+c[9,2]*t[1]^2)*(c[1,3]+c[6,3]*t[3]^2+c[3,3]*t[3])- ( c[7,3]*t[3]+c[4,3]+c[9,3]*t[3]^2)*(c[1,2]+c[2,2]*t[1]+c[4,2]*t[1]^2 ) ) : td[2]:=((c[6,2]+c[8,2]*t[1]+c[9,2]*t[1]^2)*(c[8,3]*t[3]^2+c[5,3]*t[3]+c[2,3])- ( c[7,3]*t[3]+c[4,3]+c[9,3]*t[3]^2)*(c[3,2]+c[5,2]*t[1]+c[7,2]*t[1]^2 ) ) : tn[4]:=-((c[1,4]+c[2,4]*t[3]+c[4,4]*t[3]^2)*(c[4,1]+c[7,1]*t[1]+c[9,1]*t[1]^2)- ( c[6,4]+c[8,4]*t[3]+c[9,4]*t[3]^2)*(c[1,1]+c[3,1]*t[1]+c[6,1]*t[1]^2 ) ) : td[4]:=((c[3,4]+c[5,4]*t[3]+c[7,4]*t[3]^2)*(c[4,1]+c[7,1]*t[1]+c[9,1]*t[1]^2)- ( c[6,4]+c[8,4]*t[3]+c[9,4]*t[3]^2)*(c[2,1]+c[5,1]*t[1]+c[8,1]*t[1]^2 ) ) : rbar[1]:=(c[1,1]+c[6,1]*t[1]^2+c[3,1]*t[1])*td[4]^2 + ( c[8,1]*t[1]^2+c[5,1]*t[1]+c[2,1])*td[4]*tn[4]+ ( c[7,1]*t[1]+c[4,1]+c[9,1]*t[1]^2)*tn[4]^2 : rbar[2]:=(c[1,2]+c[2,2]*t[1]+c[4,2]*t[1]^2)*td[2]^2 + ( c[3,2]+c[5,2]*t[1]+c[7,2]*t[1]^2)*td[2]*tn[2]+ ( c[6,2]+c[8,2]*t[1]+c[9,2]*t[1]^2)*tn[2]^2 : f[1]:=(x , y)->(subs({t[1]=x , t[3]=y},simplify(rbar[1 ] ) ) ) : f[2]:=(x , y)->(subs({t[1]=x , t[3]=y},simplify(rbar[2 ] ) ) ) : return(f ) : end proc ; .... _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ now we can compute resultants of @xmath174(y_0,y)$ ] and @xmath175(y_0,y)$ ] for any value @xmath176 with any precision @xmath177 , or symbolically . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .... y0:=1 ; n:=50 ; f:=twoequations(alpha , beta , gama , phi , n ) : ff[1]:=evalf[n](f[1](y0,y ) ) : ff[2]:=evalf[n](f[2](y0,y ) ) : evalf[n](resultant(ff[1]/coeftayl(ff[1],y=0,4),ff[2]/coeftayl(ff[2],y=0,4),y ) ) ; .... _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ still we do not know which of the listed 101 equations are independent , the first interesting problem in this direction is the following .
in section [ sec2 ] we described the closure of all flexible meshes @xmath118 by algebraic equations .
we conjecture that it is a complete intersection .
it is most probable that all the components of the closure are of the same dimension .
let us say a few words about the case of @xmath5-meshes .
direct calculation of the tangent space in the points of @xmath180 based on the results of the second section shows that the codimension of some of the irreducible components of @xmath180 is greater than or equal to 6 . on the other hand , the voss surfaces ( c.f .
@xcite and @xcite ) forms a set of flexible meshes of codimension 8 . in general
we suspect that all the irreducible components of @xmath180 has the same codimension between 6 and 8 .
note that for codimension 8 we know the following _ families of flexible meshes _ : * iv ) * meshes obtained from the cases i)iii ) by replacing some @xmath136 and/or @xmath187 by @xmath188 and/or @xmath189 .
for instance , if we replace @xmath190 by @xmath191 in the case of voss meshes , then we get equations @xmath192 , @xmath193 , and @xmath181 , @xmath182 for @xmath194
. * acknowledgement . *
the author is grateful to j. wallner for constant attention to this work and ch .
mller for useful remarks .
this work was performed at technische universitt graz within the framework of the project `` computational differential geometry '' ( fwf grant no .
s09209 ) .
99 v. i. arnold , _ ordinary differential equations _
( in russian ) , 2nd ed .
, `` nauka '' , moscow , 1975 ; english translation in the mit press ( 1978 ) . l. bianchi , _ sulle deformazioni infinitesime delle superficie flessibili ed inestendibili _
lincei , v. 1 ( 1892 ) , pp . 4148 .
a. i. bobenko , t. hoffmann , w. k. schief , _ on the integrability of infinitesimal and finite deformations of polyhedral surfaces _ , discrete differential geometry , a. i. bobenko , p. schrder , j. m. sullivan , g. m. ziegler , ( eds . ) , series : oberwolfach seminars , v. 38 ( 2008 ) , pp .
. h. graf , b. sauer , _ ber flchenverbiegung in analogie zur verknickung offener facettenflache _ , math .
ann . , v. 105 ( 1931 ) , pp .
a. kokotsakis , _ ber bewegliche polyeder _ , math .
, v. 107 ( 1932 ) , pp .
h. pottmann , j. wallner , _ infinitesimally flexible meshes and discrete minimal surfaces _ , monatshefte math .
, vol . 153 ( 2008 ) ,
. 347365 . | in this paper we study geometric , algebraic , and computational aspects of flexibility and infinitesimal flexibility of kokotsakis meshes .
a kokotsakis mesh is a mesh that consists of a face in the middle and a certain band of faces attached to the middle face by its perimeter . in particular any @xmath0-mesh made of quadrangles
is a kokotsakis mesh .
we express the infinitesimal flexibility condition in terms of ceva and menelaus theorems .
further we study semi - algebraic properties of the set of flexible meshes and give equations describing it . for @xmath0-meshes we obtain flexibility conditions in terms of face angles . |
the vision of automated driving systems holds a promise to change the transportation reality .
current deployments that focus on autonomous solutions pose a variety of sensors and actuators for safe driving on the road , e.g. , volvo _ drive me _ project in gothenburg and _ google car _ in california .
these autonomous solutions are based on the vehicles ability to observe obstacles in their line - of - sight .
vehicle - to - vehicle communication has the potential to improve the system confidence on the sensory information and support advanced vehicular coordination .
e.g. , when changing lanes and crossing intersections , as well as improving the road capacity by reducing the inter - vehicle distances .
however , communication failures can result in hazardous situations due to coordination based on inconsistent information shared by the participating vehicles .
consider an architecture , which figure [ fig : architecture ] ( left ) depicts , for implementing cooperative driving systems .
the communication protocol implements the mechanisms for exchanging information with other vehicles .
the control algorithm plans the vehicle motion according to the sensory information from on - board and remote sources .
note that the local control algorithms depend on the ( in general vectorial ) variable @xmath0 ( service level ) .
thus , @xmath0 is a common piece of information that all vehicles share in order to establish correct cooperation .
for instance , in vehicular platooning , @xmath0 might include the maximum acceleration levels imposed to all vehicles by the limited braking capabilities of one of them @xcite .
clearly , message loss when a new value of @xmath0 is established may lead to an inconsistent value in one or more vehicles , and thus , result in an unsafe operation of the entire cooperative system .
it is then necessary to have an additional layer , shown in figure [ fig : architecture ] right , between the communication layer and the control algorithm .
we propose to base this additional layer on a timed protocol for cooperation with disagreement correction that resolves disagreements on variable @xmath0 among the system vehicles . specifically , we address the following research question : how can cooperative systems be used to attain the highest performance without compromising safety in the presence of communication failures ?
we consider applications in which the individual vehicles estimate their ability to cooperate according to the sensory information quality and communicate their maximum supported cooperative level @xcite .
the vehicular system then decides on its cooperative service level according only to the received information .
however , communication failures can cause the arrival of the needed information not to occur by the deadline .
this can bring the vehicles to operate at distinct levels .
it is a critical issue to guarantee that the uncertainty period along vehicles occurs only in short time periods .
therefore , we address problem [ pr:1 ] .
[ pr:1 ] is there an upper - bound on the longest period in which the cooperative system may have inconsistent operation service level ?
we note that we can not solve problem [ pr:1 ] using distributed ( uniform ) consensus algorithms . in the uniform consensus problem ,
every component ( vehicle ) proposes a value and the objective is to select exactly on of these proposed values .
it is well - known that this problem is not deterministically solvable in unreliable synchronous networks and any @xmath1-communication rounds algorithm has probability of disagreement of at least @xmath2 @xcite ( theorem 5.1 and theorem 5.5 ) .
therefore , when the communication failures are too frequent and severe , the uncertainty period can not be bounded since the components ( vehicles ) can disagree for an unbounded number of protocol executions .
this work presents a communication protocol that guarantees the shortest possible uncertainty period , i.e. , a constant time , in the presence of communication failures .
our solution is based on a communication protocol that collects values from all system components . once this
proposed set @xmath0 is delivered to all the components , the protocol employs a deterministic function to decide on a single value from @xmath0 that all system components are to use .
the protocol identifies the periods in which there is a clear risk for disagreement due to temporary communication failures , i.e. , a period in which @xmath0 was not delivered by the due time to the entire system .
once such risk is identified , the protocol triggers a correction strategy against the risk of having disagreement for more than a constant number of rounds .
namely , after the occurrence of communication failures that jeopardize safety , all system components will rapidly start a period to reestablish their confidence by returning to the default value .
once the network returns to be stable again , and no communication failures occur , within a constant time , the protocol behaves as if no communication failures has ever occurred . the correctness proof and its validation show that the proposed solution provides a trade - off between the uncertainty period ( in the order of milliseconds ) and the occurrence of communication failures . in other words , for shorter round length ( and consequently so it the uncertainty period ) , the vehicles experience more frequently a low service level .
however , for a longer round length , the vehicles experience less frequently a low service level . however , the longer the round length is , the longer the time that vehicles spend on disagreements and therefore , the risk of having accidents increases .
this paper also discusses a safety - critical application that facilitates cooperation using the proposed protocol .
we assume a baseline adaptive cruise control ( acc ) that does not require communication .
then , we extend it to a cooperative one that attains higher vehicle performance , but relies on higher confidence level about the position and velocity of nearby vehicles .
we explain how the protocol can provide a timed and distributed mechanism for facilitating decisions about when the vehicles should plan their trajectories according to the baseline application and when according to the extended one that fully utilizes cooperative functionality .
the distributed ( uniform ) consensus problem considers the selection of a single value from a set of values proposed by members of , say , a vehicular system . the solution is required to terminate within a bounded time by which all system components have decided on a common value .
the use of the exact ( uniform ) vs. approximate consensus approaches is explain here @xcite , where they recommend the use of exact ( uniform ) consensus due to the simplicity of the system design from the application programmer perspective .
the exact consensus approach , in contrast to the approximate one , rests on a foundation of clearly defined requirements and is amenable to formal methods and analytical validation .
a number of impossibility results consider distributed consensus in general ( see @xcite ) . in @xcite
, the author shows that the presence of communication failures makes impossible to deterministically reach consensus ( theorem 5.1 ) and any @xmath1-round algorithm has probability of disagreement of at least @xmath2 ( theorem 5.5 ) .
this implies that there are no guarantees that vehicles can reach consensus on bounded time since vehicle - to - vehicle communications are prone to failures .
moreover , when the communication failures are too frequent and severe , vehicles can fail to reach consensus for an unbounded number of consecutive times .
we therefore abandon consensus - based decision algorithms , and prefer to focus on solutions that offer early fall - back strategies against the risk of having disagreement for more than a constant number of rounds .
the existing literature on distributed ( uniform ) consensus algorithms with real - time requirements does consider processor failures .
however , it often assumes timed and reliable communication .
for example , in @xcite the authors give an algorithm that reaches agreement in the worst case in time that is sublinear in the number of processors and maximum message delay . in @xcite ,
the authors provide a time optimal consensus algorithm that reaches consensus in time @xmath3 in the worst case where @xmath4 is the maximum message delay and @xmath5 the maximum number of processors that can crash . in this paper , we do not assume reliable communication .
thus , message drops can occur independently among processors at any time .
group communication systems @xcite treat a group of participants as a single communication endpoint .
the group membership service @xcite monitors the set of recently live and connected participating system components whereas the multicast service delivers messages to that group under some delivery guarantees , such as delivery acknowledgment . in this paper
we assume the existence of a membership service and a best - effort ( single round solution ) dissemination ( multicast ) protocol that has no delivery acknowledgment .
there exists literature on adaptive cruise control @xcite as well as vehicle platooning @xcite . in @xcite ,
the author considers vehicle platooning and lane merging , and bases his construction on distributed high level communication primitives .
we consider a different failure model for which there is no deterministic implementation for these communication primitives .
the studied problem is motivated by the karyon project @xcite .
the karyon project aims to provide a predictable and safe coordination of intelligent vehicles that interact in inherently uncertain environments .
it proposes the use of a safety kernel that enforces the service level that the vehicle can safely operate .
a cooperative service level can ensure that vehicles follow the same performance level . in this paper , we study a communication protocol that implements the karyon s cooperative service level evaluator . in @xcite
, we present the architecture that considers the interactions between the safety kernel , a local dynamic map and the cooperative service level evaluator .
unlike the earlier abstract presentation of the cooperative service level evaluator , this paper provides in detail , the design and analysis of the communication protocol .
we study an elegant solution for cooperative vehicular systems that have to deal with communication uncertainties .
we base the solution on a communication protocol that , we believe , can be well understood by designers of safety - critical , automated and cyber - physical systems .
we explain how the designers of fault - tolerant cooperative applications can use this solution to deal with communication failures when uniformly deciding on a shared value , such as @xmath0 .
we consider cooperative applications that must periodically decide on a shared values @xmath0 .
since the consensus problem can not be deterministically solved in the presence of communication failures , the system is doomed to disagreed on the value of @xmath0 ( in the presence communication failures that are frequent and severe ) .
we bound the period in which the vehicles can be unaware of such disagreements with respect to @xmath0 . we prove and validate that this bound is no more than one communication round ( in a vehicular system that deploys a single - hop network of wireless ad hoc communication ) .
we also study the percentage of time during which the system avoids disagreement on @xmath0 using ns-3 simulations .
we exemplify how the proposed solution helps to guarantee safety .
we consider vehicles that operate in a cooperative operational mode as long as they are aware that all the nearby vehicles are also in the same mode ( with at most one communication round period of disagreement ) .
however , if at least one vehicle is suspecting that another vehicle is not , all vehicles switch , within one communication round period , to a baseline operational mode so that the safety standards are met .
we list our assumptions and define the problem statement ( section [ s : sys ] ) , before providing the timed protocol for cooperation with disagreement correction ( section [ s : dcp ] ) and its correctness proof ( section [ s : c ] ) . as protocol validation study , we consider computer simulation ( section [ s : eva ] ) . we discuss cooperative vehicular application ( section [ s : cva ] ) and an example before the conclusions ( section [ s : con ] )
[ s : sys ] we consider a message passing system that includes a set @xmath6 of @xmath7 communicating prone - resilient vehicles .
we refer to the vehicles with i d @xmath8 as @xmath9 .
we assume that all vehicles have access to a common global clock with a sub - microsecond offsets by calling the function @xmath10 .
this could be implemented , for example , using global positioning systems ( gps ) @xcite .
hence , we assume that the maximum time difference along vehicles is at most @xmath11 .
we consider that the system runs on top of a timed and fault - tolerant , yet unreliable , dissemination protocol , such as @xcite , that uses @xmath12 to broadcast message @xmath13 from vehicle @xmath14 to all vehicles in @xmath6 .
we assume that end - to - end message delay is at most @xmath15 time .
thus , messages are either delivered within @xmath15 time or omitted . the constant @xmath15 depends on distinct factors such as the mac protocol that is used , vehicle velocity , interference , etc .
for example , this bound can be set to @xmath16ms or less using , for example , dedicated short - range communications ( dsrc ) @xcite .
vehicle @xmath17 receives @xmath13 from @xmath8 by raising the event @xmath18 .
we consider a fully connected network topology .
however , the network can arbitrarily decide to omit messages , but not to delay them for more than @xmath15 time .
these assumptions allow the protocol to run in a synchronous round based fashion .
we consider rounds of time @xmath19 where @xmath20 .
every vehicle @xmath9 executes a program that is a sequence of _ ( atomic ) steps_. an input event can be either the receipt of a message or a periodic timer going off triggering @xmath9 to start a new iteration of the do forever loop .
we define the _ uncertainty period _ as the period that vehicles can disagree .
we say that there was a _ communication failure _ at round @xmath1 if there exists a vehicle that has not received the messages from all vehicles during round @xmath1 .
the system s task is to satisfy requirements [ prp : dis1 ] to [ prp : dis3 ] , which consider definition [ def : scp ] .
[ def : scp ] a stable communication period @xmath21 $ ] is the period of @xmath22 rounds in which the vehicles do not experience communication failures , i.e. , all vehicle receive all messages during these periods .
otherwise , it is called an unstable communication period , denoted by @xmath23 $ ] .
we say that a stable communication period @xmath21 $ ] is maximal when rounds @xmath24 and @xmath25 are unstable communication periods .
analogously , we define a maximal unstable communication period @xmath23 $ ] , see figure [ fig : be ] .
thus , in any run , the communication may go through maximal stable and unstable periods ( and then perhaps back to stable ) for an unbounded number of times .
requirements [ prp : dis1 ] to [ prp : dis3 ] deal with what the system output at every vehicle should be when it goes between the different periods . $ ] followed by a maximal stable communication period @xmath26 $ ] . ]
[ prp : dis1 ] during a stable period no two vehicles use different values .
moreover , within a bounded prefix of every stable period , there is a suffix during which no uses the default return value .
[ prp : dis2 ] every unstable period has a suffix named the _ disagreement correction period _ during which no two vehicles use different values . during this period
all vehicles use the default return value .
[ prp : dis3 ] the suffix of a stable period during which some vehicles may use different values is called the _
uncertainty period_. we require it to be bounded .
we show that any system run of the proposed solution fulfills requirements [ prp : dis1 ] to [ prp : dis3 ] .
specifically , we demonstrate theorem [ th : main ] ( section [ s : dcp ] ) .
[ th : main ] the proposed protocol ( algorithm [ alg : tddp ] ) fulfills requirements [ prp : dis1 ] , [ prp : dis2 ] and [ prp : dis3 ] , where the uncertainty period is bounded by one round .
moreover , if vehicles do no experience communication failures , the disagreement correction holds for at most one round .
[ s : dcp ] we present the communication protocol in which the participants exchange messages until a deadline .
these messages can include information , for example , about nearby vehicles as well as the confidences that each vehicle has about its information .
once everybody receives the needed information from each other , the participants can locally and deterministically decide on their actions . in case of a communication failure , each participant that experiences a failure imposes the default return value for one round .
each vehicle @xmath14 executes the protocol ( that algorithm [ alg : tddp ] presents ) .
it uses a do forever loop for implementing a round base solution .
it accesses the global clock ( line [ ln : clock ] ) and checks whether it is time for the vehicle to send information about the current round ( line [ ln : roundclock1 ] ) .
a vehicle starts sending messages at @xmath11 time from the beginning of each round and @xmath27 before of the end of each round using the @xmath28 interface ( line [ ln : gossipcall ] ) .
recall that @xmath11 is the maximum time difference over the vehicles and @xmath15 is the longest time that a message can live in the network .
next , it tests whether the current round number @xmath29 points to the current round in time ( line [ ln : roundclock ] ) .
a new round starts when @xmath30 is greater than @xmath29 . at the beginning of every round ,
the protocol first keeps a copy of the collected data and the received information , and updates the round counter , as well as nullifying @xmath31 and @xmath32 ( line [ ln : roundnull ] ) .
then , it tests whether it has received all the needed information for the previous round ( line [ ln : arrivaltest ] ) .
suppose that a communication failure occurred in the previous round , the protocol sets the data to be sent to the default return value @xmath33 ( line [ ln : assingment1 ] ) .
it also writes to @xmath34 interface the received information as well as the default return value @xmath33 ( line [ ln : writeoutput1 ] ) .
however , in the case that all messages of the previous round have arrived on time , the system reads the application information using @xmath35 interface .
it also writes to @xmath34 interface the received information as well as the value that the deterministic function @xmath36 returns ( line [ ln : imposeend ] ) . the proposed protocol interfaces with the gossip ( dissemination ) protocol by sending messages ( @xmath28 ) and receiving them ( @xmath37 ) periodically .
the protocol locally stores the arriving information from @xmath38 on each round in @xmath39 $ ] and waits for the round end before it finishes accumulating all arriving information . more specifically , for each message that is reported with the same round , the protocol stores the data from @xmath40 and sets the acknowledgment variable to true , if the message comes directly from @xmath17 , ( @xmath41 ) , or transitively from @xmath17 without considering its own values ( @xmath42 $ ] and @xmath43 ) .
the correctness proof shows that , in the presence of a single communication failure , there could be at most one disagree round in which different system components use different values .
moreover , the influence of that single failure will last for at most two rounds , which is the shortest period possible .
note that algorithm [ alg : tddp ] handles well any sequence of communication failures .
[ bt ] @xmath33 : denotes a void ( initialized ) entry , as well as the default return value .
@xmath11 : the maximum time difference among vehicle clocks .
@xmath44 : the maximum time that a message time can live in the network .
@xmath45 : the length of a round .
+ @xmath46 : current communication round .
@xmath47 : current clock .
@xmath48 = \{\ldots \}$ ] : application data where @xmath39 $ ] is the data received at round @xmath29 from member @xmath40 .
@xmath49 = \{false , \ldots \}$ ] : acknowledge for data reception where @xmath50 $ ] is true if @xmath9 has received ( directly or indirectly ) the message from @xmath40 of the current round .
+ @xmath28 : disseminate information to the system members .
@xmath37 : dispatch arriving messages .
@xmath35 : return a datum to be sent .
@xmath34 : write decided output . @xmath51 deterministically determines an item from @xmath0 .
we assume that whenever @xmath52 , then @xmath53 .
[ al : fun ] + + @xmath54 [ ln : gossipreceive ] @xmath55,ack[k ] ) \gets ( data_j[k ] , true)$ ] @xmath56 [ ln : clock ] [ ln : roundclock1 ] @xmath57[ln : gossipcall ] [ ln : roundclock ] @xmath58 ) \gets ( data , ack , myclock \div roundlength , \{\bot , \ldots \ } , k = i : \forall p_k \in members)$][ln : roundnull ] [ ln : arrivaltest ] @xmath59 \gets \bot$ ] [ ln : assingment1 ] @xmath60 [ ln : writeoutput1 ] @xmath59 \gets readstate()$ ] [ ln : assingment3 ] @xmath61 [ ln : imposeend ]
[ s : c ] we prove that algorithm [ alg : tddp ] follows requirements [ prp : dis1 ] to [ prp : dis3 ] .
[ th : boundeduncertainty ] let @xmath62 $ ] be any maximal unstable communication period followed by a maximal stable communication period @xmath63 $ ] .
the following three statements hold .
\(1 ) * bounded uncertainty period*. vehicles may have disagreements at round @xmath64 .
\(2 ) * disagreement correction*. all vehicles use the default return value during @xmath65 $ ] .
\(3 ) * certainty period*. vehicles use the same value during @xmath66 $ ] .
let @xmath67 be the set of messages that vehicle @xmath9 receives from all the vehicles @xmath68 , either directly or indirectly , and that @xmath17 has sent during round @xmath1 .
observe that each vehicle decides the value to be used on round @xmath69 based on the received information at round @xmath1 ( lines [ ln : writeoutput1 ] and [ ln : imposeend ] ) .
we claim that @xmath70 for @xmath71 and @xmath72 $ ] .
note that this implies that no two vehicles use different values when processing round @xmath1 , because vehicle @xmath9 determines its output value according to the deterministic function @xmath73 .
@xmath74 for @xmath71 where @xmath75 $ ] .
first we show that each vehicle maintains consistent its own information over each round .
observe that lines between [ ln : roundnull ] and [ ln : imposeend ] are executed once during round @xmath29 since @xmath29 is set to @xmath76 and @xmath10 always returns larger values .
therefore , each vehicle @xmath9 loads its message on the register @xmath59 $ ] once during @xmath29 .
thus , assume that vehicle @xmath77 overwrites its @xmath59 $ ] when receiving a message from vehicle @xmath78 .
since the condition ensures that it loads @xmath39 $ ] only if either @xmath79 or @xmath80 , we conclude that @xmath81 .
thus , @xmath59 $ ] is consistent on @xmath9 during round @xmath29 .
we say that a message @xmath82 is sent transitively , if @xmath9 receives @xmath82 from @xmath17 where @xmath83 .
we show that the message transitivity maintains the consistency of the messages during a stable communication period .
we argue by contradiction .
assume that there are two messages , @xmath84 and @xmath85 such that @xmath86 .
consider the first time that @xmath87 were sent .
observe that @xmath9 sent the two messages . a contradiction
since @xmath9 maintains consistent its own information over each round .
the claim follows by showing that at the end of the current round @xmath29 , it holds that @xmath88 . indeed ,
since messages of each round are sent @xmath89 time units before the end of @xmath29 and after @xmath11 time units after the beginning of @xmath29 , vehicles receive messages only from the current round .
recall that @xmath11 is the maximum difference time among vehicle clocks and @xmath44 is the maximum time that a message can live in the network .
\(1 ) * bounded uncertainty period*. consider round @xmath90 .
since @xmath62 $ ] is an unstable communication period , there exists a vehicle @xmath9 that did not receive a message from all vehicles , i.e. , @xmath33 is in @xmath32 .
observe that vehicle @xmath17 is unaware that @xmath9 had experienced a communication failure during round @xmath90 .
let us assume that @xmath17 did not experience any communication failure .
therefore , @xmath17 uses the deterministic value that @xmath36 returns on round @xmath64 .
however , @xmath9 imposes the default return value ( line [ ln : writeoutput1 ] ) since it had experienced a communication failure .
thus , during round @xmath64 , @xmath9 sends the default return value by setting @xmath59 $ ] to @xmath91 and uses it ( lines [ ln : assingment1 ] and [ ln : writeoutput1 ] , respectively ) .
therefore , as long as no vehicle misses @xmath9 s message , the first default return value of @xmath9 arrives along round @xmath92 .
thus , during round @xmath64 , @xmath17 uses a distinct value than @xmath9 .
\(2 ) * disagreement correction*. we show that all vehicles use the default return value in round @xmath93 $ ] .
it is sufficient to show that there exists at least one default return value in @xmath94 in each round @xmath95 $ ] .
assume that at round @xmath96 $ ] , some vehicle @xmath40 experienced a communication failure . therefore , at round @xmath69 all other vehicles either receive the default value of @xmath40 or receive no message from @xmath40 .
thus , all vehicles use the same value ( default return value ) in each round @xmath93 $ ] ( lines [ ln : assingment1 ] and [ ln : imposeend ] ) .
this is due to the definition of the function @xmath36 ( line [ al : fun ] ) and the fact that each vehicle writes the default return value if it experiences a communication failure .
\(3 ) * certainty period*. we show that during @xmath66 $ ] all vehicles use the same values .
indeed , from the point ( 2 ) , every vehicle uses the default return value in every round @xmath93 $ ] . it remains to show that they use the same value in each round @xmath97 $ ] .
from the claim , @xmath98 for each pair @xmath99 during @xmath100 $ ] since all vehicles received the information from each other vehicle .
the lemma follows since vehicles decide the value to be used on round @xmath69 based on the received information at round @xmath1 ( lines [ ln : writeoutput1 ] and [ ln : imposeend ] ) using the deterministic function @xmath36 .
it follows directly from lemma [ th : boundeduncertainty ] .
[ s : eva ] we consider a cooperative system that has two service levels where the lowest one is the default service level to which the system falls - back to in the presence of communication failures .
for example , this can be a vehicular system in which the cooperative service level is the highest , and the autonomous service level is the lowest ( default ) one .
since we focus on communication failures , the experiments assume that every system component can always support the highest service level , and thus read input ( @xmath101 ) always returns the highest service level .
we use computer simulation to validate the protocol as well as its efficiency . for the efficiency , we consider the _ reliability _ performance measure which we define as the percentage of communication rounds during which the protocol allows the system to run at its highest service level .
first , we validate that the disagreement period is of at most one round and next the reliability of the protocol .
we simulate the protocol using ns-3 .
we choose ieee 802.11p as the communication channel with a log - distance path loss model and nakagami fading channel model .
since dsrc technologies support end - to - end message delay of less than @xmath16ms @xcite , we fix the message delay to @xmath16ms .
we consider a synchrony bound of @xmath102ms , say , using gps @xcite or a distributed clock - synchronization protocol .
we implement a straightforward gossip protocol in which every node retransmits message every @xmath103ms .
we validate that the disagreement period is of at most one round .
we plot in figure [ fig : be ] the decision that @xmath104 vehicles took independently during @xmath105 rounds using the protocol under frequent communication failures .
we set the round length to @xmath106ms so that messages can be transmitted twice in each round .
observe that at round @xmath107 vehicles @xmath108 and @xmath109 reduce the service level due to a communication failure , but vehicles @xmath110 and @xmath104 still continue in the highest level of service .
however , at round @xmath111 , they lower their service level .
although vehicles do not operate on distinct service levels for more than one round , the service level of some vehicles may be oscillating .
we can reduce this effect by increasing the round length . however , the uncertainty period also increases .
rounds using the protocol . ]
note the trade - off between the upper bound on the disagreement period , which is one communication round , and the success rate of the gossip protocol , which decreases as the round length becomes shorter .
the type of gossip protocol as well as the number of system components also influences this success rate .
we use computer simulation to study how these trade - offs work together and present the reliability .
we consider three round lengths between @xmath106ms and @xmath112ms with intervals of @xmath16ms so that vehicles can transmit @xmath113 and @xmath114 messages in each round , respectively .
we variate the number of vehicles between two and eight .
the reliability of the system is plotted in figure [ fig : rel ] .
we run each experiment for @xmath112 simulation seconds . during the simulations ,
we observe a packet drop average of @xmath115 .
the packet drop rate per number of vehicles is presented in table [ fig : packetdrop ] .
further , the percentage of time that all vehicles agree on the highest service level is greater than @xmath116 with round lengths of at least @xmath117ms with at least four vehicle .
observe that the reliability is higher with more vehicles than with less .
this is because of the transitivity property . .packet drop rate . [
cols="^,^",options="header " , ] the protocol that algorithm [ alg : tddp ] presents . @xmath51 @xmath118.locallos)$ ] + + @xmath35 let @xmath119 and @xmath120 be the operation information and maximum local service level that it supports , respectively , of @xmath9 @xmath121 + + @xmath122 use @xmath31 and acceleration bounds provided in table [ fig : saftykernelparameters1 ] according to @xmath123 to maintain the cruise velocity if possible use @xmath31 and acceleration bounds provided in table [ fig : saftykernelparameters1 ] according to @xmath123 to maintain the headway given in table [ fig : saftykernelparameters ] .
algorithm [ alg : platoonadaptivecruisecontrol ] executes an instance of the protocol ( that algorithm [ alg : tddp ] presents ) and implement the interface functions @xmath101 , @xmath124 and @xmath125 . the function
@xmath101 returns the @xmath9 s local service level ( see table [ fig : saftykernelparameters1 ] ) as well as the operational information ( localization , heading , velocity , etc . ) .
the function @xmath125 returns the minimum local service level in the data structure @xmath0 so that all vehicles can meet the required constrains .
the main functionality is implemented in @xmath124 function .
it uses the information of all the vehicle in @xmath31 to determine the velocity and acceleration for the next round according to the cooperative service level using the parameters in table [ fig : saftykernelparameters ] . note that in the baseline application , acc , vehicles can base their decision on sensory information from onboard sources .
we assume that each operation mode is proven to be safe provided that the information meets the requirements , i.e. , the errors are within the bounds that are given in table [ fig : saftykernelparameters ] . for the worst case scenario ,
the behavior of the platoon can be influenced by vehicles that are not part of the platoon .
this is because some events can cause cascade effects if they occur during the communication failures .
we observe that the period during which the system switches from the highest service level to the lowest is a critical time .
the safety provision in algorithm [ alg : platoonadaptivecruisecontrol ] depends directly on the mechanical constraints and the parameters election . from the previous section ,
it is reasonable to consider rounds of length at least @xmath117 milliseconds .
thus , the headway can be determined from the round length and the error bounds on the information .
we observe that algorithm [ alg : platoonadaptivecruisecontrol ] can reduce the collision risk by enforcing the vehicles to operate in a common service level that has been proven to be safe according to the information quality that is associated with that level . as an illustrative example of the propose solution
, we consider three vehicles in the following worst case scenarios of two implementations : ( 1 ) a vehicular platooning application that does use the proposed solution for for its back - off strategy in the presence of communication failures , and ( 2 ) an implementation of vehicular platooning that does run algorithm [ alg : platoonadaptivecruisecontrol ] .
let @xmath126 be the vehicles such that @xmath127 is leading the platoon followed by @xmath128 and @xmath129 as depicted in figure [ fig : exm1 ] .
assume that vehicles are driving on platooning with operational parameters given by the medium service level in table [ fig : saftykernelparameters ] .
therefore , they keep a headway of @xmath130 .
suppose that at time @xmath131 , @xmath128 starts loosing the messages from @xmath129 for @xmath132 time .
further , assume that at time @xmath133 , @xmath128 losses the messages from @xmath127 for @xmath134 time and at the same time @xmath127 requires to decelerate due to an obstacle , for example a pedestrian .
let us assume that @xmath135 and @xmath132 are at least two times the round length .
* platooning with back - off strategy that does not include the proposed solutions . *
since @xmath128 does not receive the messages from @xmath129 during @xmath136 $ ] , it is unaware whether @xmath129 continues operating on platooning .
thus , @xmath128 continues operating on platooning and assumes that it is the last vehicle in it . at time @xmath133
, @xmath128 starts loosing messages from @xmath127 and consequently switches to the back - off strategy in the next round .
however , since @xmath127 requires to brake , @xmath128 uses the acceleration bounds in @xmath137 . but @xmath129 continues operating on platooning during @xmath136 $ ] , since it is unaware that @xmath128 is not receiving messages from @xmath127 and @xmath129 . by definition ,
the system is not safe during @xmath138 $ ] , since the platoon has only been proved to be safe when the headway is at most @xmath130 and acceleration bounds are in @xmath139 . *
platooning using algorithm [ alg : platoonadaptivecruisecontrol ] .
* from the algorithm property that the uncertainty does not hold for more than one round , @xmath127 and @xmath129 will be aware that at least one vehicle has a communication failure in the next round .
therefore , all switch to the lowest service level and start opening space to keep a headway of @xmath140 .
thus , at time @xmath133 they have larger inter - vehicle distances which reduce the cascade effects .
observe that for an @xmath135 less than two round lengths , the problem also occurs in this approach .
indeed , every cooperative vehicular application that relies on communication suffers from this problem .
however , we believe that our approach minimizes the effects .
[ s : con ] we have proposed an efficient protocol that can be used in safety - critical cooperative vehicular applications that have to deal with communication uncertainties .
the protocol guarantees that all vehicles will not be exposed , for more than a constant time , to risks that are due to communication failures .
we demonstrate correctness , evaluate performance and validate our results via ns-3 simulations .
we also showed how vehicular platooning can use the protocol for maintaining system safety .
the proposed solution can be also extended to other cooperative vehicular applications , such as intersection crossing , coordinated lane change , as we demonstrated using the gulliver test - bed @xcite during the karyon project @xcite .
moreover , we have considered the simplest multi - hop communication primitive , i.e. , gossip with constant retransmissions . however , that communication primitive can be substitute with a gossip protocol that facilitate a greater degree of fault - tolerance and better performance .
this work opens the door for the algorithmic design and safety analysis of many cooperative applications that use different high - level communication primitives .
marcos kawazoe aguilera , grard le lann , and sam toueg . on the impact of
fast failure detectors on real - time fault - tolerant systems . in dahlia
malkhi , editor , _ disc _ , volume 2508 of _ lecture notes in computer science _ , pages 354370 .
springer , 2002 .
christian berger , oscar morales ponce , thomas petig , and elad michael schiller . driving with confidence : local dynamic maps that provide los for the gulliver test - bed . in andrea bondavalli , andrea ceccarelli , and frank ortmeier , editors ,
_ computer safety , reliability , and security - safecomp 2014 workshops : ascoms , decsos , devvarts , isse , resa4ci , sassur .
florence , italy , september 8 - 9 , 2014 .
proceedings _ , volume 8696 of _ lecture notes in computer science _ , pages 3645 .
springer , 2014 .
antonio casimiro , jrg kaiser , johan karlsson , elad michael schiller , philippas tsigas , pedro costa , jos parizi , rolf johansson , and renato librino .
brief announcement : karyon : towards safety kernels for cooperative vehicular systems . in andra
w. richa and christian scheideler , editors , _ sss _ , volume 7596 of _ lncs _ , pages 232235 .
springer , 2012 .
antonio casimiro , oscar morales ponce , thomas petig , and elad michael schiller .
vehicular coordination via a safety kernel in the gulliver test - bed . in _
34th international conference on distributed computing systems workshops ( icdcs 2014 workshops ) , madrid , spain , june 30 - july 3 , 2014 _ , pages 167176 .
ieee , 2014 .
antnio casimiro , jos rufino , ricardo c. pinto , eric vial , elad m. schiller , oscar morales - ponce , and thomas petig . a kernel - based architecture for safe cooperative vehicular functions . in
9th ieee international symposium on industrial embedded systems ( sies14 ) _ , 2014 .
se lee , e llaneras , s klauer , and j sudweeks .
analyses of rear - end crashes and near - crashes in the 100-car naturalistic driving study to support rear - signaling countermeasure development . , 810:846 , 2007 .
oscar morales - ponce , elad m. schiller , and paolo falcone .
cooperation with disagreement correction in the presence of communication failures . in _
intelligent transportation systems ( itsc ) , 2014 ieee 17th international conference on _ , pages 11051110 , oct 2014 .
mitra pahlavan , marina papatriantafilou , and elad michael schiller .
gulliver : a test - bed for developing , demonstrating and prototyping vehicular systems .
in jos d. p. rolim , jun luo , and sotiris e. nikoletseas , editors , _ proceedings of the 9th acm international workshop on mobility management & wireless access , mobiwac 2011 , october
31- november 4 , 2011 , miami beach , fl , usa _ , pages 18 .
acm , 2011 .
mitra pahlavan , marina papatriantafilou , and elad michael schiller .
gulliver : a test - bed for developing , demonstrating and prototyping vehicular systems . in _ proceedings of the 75th ieee
vehicular technology conference , vtc spring 2012 , yokohama , japan , may 6 - 9 , 2012 _ , pages 12 .
ieee , 2012 .
steven e shladover , charles a desoer , j karl hedrick , masayoshi tomizuka , jean walrand , w - b zhang , donn h mcmahon , huei peng , shahab sheikholeslam , and nick mckeown .
automated vehicle control developments in the path program .
, 40(1):114130 , 1991 . | vehicle - to - vehicle communication is a fundamental requirement for maintaining safety standards in high - performance cooperative vehicular systems .
the vehicles periodically exchange critical information among nearby vehicles and determine their maneuvers according to the information quality and the established strategies
. however , wireless communication is failure prone .
thus , participants can be unaware that other participants have not received the needed information on time .
this can result in conflicting ( unsafe ) trajectories .
we present a deterministic solution that allows all participants to use a fallback strategy in the presence of communication delays .
we base our solution on a timed distributed protocol . in the presence of message omission and delay failures ,
the protocol disagreement period is bounded by a constant ( in the order of milliseconds ) that may depend on the message delay in the absence of these failures .
we demonstrate correctness and perform experiments to corroborate its efficiency .
we explain how vehicular platooning can use the proposed solution for providing high performance while meeting the safety standards in the presence of communication failures .
we believe that this work facilitates the implementation of cooperative driving systems that have to deal with inherent ( communication ) uncertainties . |
there is presently a lot of interest in using linear alkylbenzene ( lab ) as a scintillator solvent for neutrino experiments ( _ e.g. _ sno , lena , reno and daya bay @xcite ) .
first proposed as a scintillator by m. chen , this organic liquid has the advantages of low toxicity , good light yield , high flash - point and long attenuation length .
furthermore , it is compatible with acrylic , a material that is frequently used to build the scintillator containment vessel . to understand the propagation of light and
reconstruct events in these experiments , and to accurately calculate the cherenkov light production , it is imperative to know the refractive index of the liquid scintillator as a function of wavelength
. currently , our knowledge of the refractive index of lab - based scintillators is limited to the optical region .
recently , the reno collaboration reported a measurement of the refractive index of lab at six wavelengths in the interval @xmath3 nm , using the mininum deviation technique @xcite . in this paper
, we describe a measurement of the real component of the refractive index , @xmath4 , of two lab - based scintillators , to the nearest @xmath0 at 500 points from 210 to 1000 nm , using an ellipsometric method @xcite .
these two scintillators were : lab with 3 g / l of 2,5-diphenyloxazole ( ppo ) , and lab - ppo loaded with 0.094 % nd by weight .
we also report on a measurement of @xmath4 for ej-301 scintillator @xcite , which is equivalent to the very commonly - used ne-213 and bc-501a scintillators .
as far as we know , the dispersion of ne-213 has not been measured down to the uv region .
the use of ellipsometers for measuring the refractive indices of liquids is convenient , but not widespread . because of challenges brought about by surface vibrations , evaporation and difficulties in controlling the temperature of the liquid medium , the ellipsometric method is not expected to be as competitive as the minimum deviation technique for precision refractive index measurements .
however , measurements that are accurate to the third decimal place are possible . in this work
, we adopted the simplest ellipsometric configuration , in which polarized light reflects off an infinitely deep layer of liquid .
this is described below . .
( b ) set - up with liquid film on a frosted glass slide .
this is equivalent to a deep container .
, title="fig:",width=291 ] .
( b ) set - up with liquid film on a frosted glass slide .
this is equivalent to a deep container .
, title="fig:",width=291 ] consider polarized light incident on the interface between two media , both of which are infinite and homogeneous ( figure [ figure : reflectionfigures](a ) ) .
after reflection , there is a change in polarization that is dependent on the incident angle , the refractive indices of the two media , as well as the initial polarization state .
the ellipsometer measures the change in polarization in terms of two parameters , @xmath5 and @xmath6 .
if @xmath7 and @xmath8 are the reflection coefficients of the electric field components parallel ( @xmath9 ) and perpendicular ( @xmath10 ) to the plane of incidence , the ellipsometric parameters are given by : @xmath11 @xmath7 and @xmath8 are expressed in terms of the complex refractive indices of media 1 and 2 ( @xmath12 and @xmath13 ) , and the angles @xmath14 and @xmath15 ( figure [ figure : reflectionfigures](a ) ) by fresnel s equations . combining fresnel s equations and snell s law with equation [ equation : ellipsometer ] , and separating real and imaginary components , one gets @xcite : @xmath16=x\ ] ] @xmath17 where @xmath18 and @xmath19 are respectively , the real components of @xmath12 and @xmath13 , and @xmath20 is the imaginary component of @xmath13 . given @xmath5 , @xmath6 , @xmath14 and @xmath18
, one can therefore solve for @xmath19 and @xmath20 separately : @xmath21 in our case , medium 1 is air , and medium 2 is the liquid scintillator , whose real refractive index component ( @xmath19 ) is to be measured .
the sensitivity to @xmath20 is weak , since the measurement is made on the reflected , and not the transmitted light component , which conveys most of the information on attenuation in medium 2 .
for equations [ equation : n21 ] and [ equation : n22 ] to be valid , one must have : ( 1 ) an ideal interface , and ( 2 ) an infinitely deep medium 2 ( so that no interference occurs with light reflected off the bottom of medium 2 ) . to emulate a deep container with a flat liquid surface
, we followed synowicki _
@xcite : a few drops of liquid were allowed to spread into a thin film , on a sandblast - frosted microscope glass slide ( figure [ figure : reflectionfigures ] ( b ) ) .
measurements were then made with the light reflected specularly off the upper surface of the liquid film .
the transmitted component was diffusely reflected off the frosted glass and did not affect the results .
the slides , as well as all other glassware used in this work , were first cleaned in an ultrasound bath .
data was taken at five values of @xmath14 , namely @xmath22 and @xmath23 .
the experiment was performed at room temperature ( 25@xmath24 c ) with a variable - angle woollam m2000 ellipsometer , fitted with a xe lamp . excluding alignment procedures , a spectral scan at one angle typically took around 30 s.
results for lab - ppo are shown in figure [ figure : labresults](a ) .
the measurements at the five different angles agree within around @xmath0 over 2001000 nm .
the solid curve is the average from the five angles , while the shaded region shows the standard deviation . shown in the circle markers are measurements from reno @xcite , obtained using the mininum deviation technique .
the diamond markers are previous measurements by the sno collaboration @xcite , taken using an abbe refractometer and other methods .
the nd - doped lab - ppo results are shown in figure [ figure : labresults](b ) .
the values are rather close to undoped lab - ppo , the difference being within @xmath250.5 % over the entire wavelength range under study . for the wavelength range 2301000 nm ,
where the refractive indices of both liquids decrease monotonically with wavelength @xmath26 , we parameterize the dispersion with the sellmeier formula : @xmath27 with @xmath28 . for the range 210230 nm
, we fit to a third order polynomial : @xmath29 in both equations [ equation : sellmeier ] and [ equation : polynomial ] , @xmath26 is in nm . the sellmeier and polynomial coefficients are shown in tables [ table : sellmeiercoeff ] and [ table : polycoeff ] , respectively .
the fit residuals are shown in figure [ figure : fitresiduals ] .
@ccccccccc liquid&@xmath30&@xmath31&@xmath32&@xmath33&@xmath34&@xmath35&@xmath36&@xmath37 + lab - ppo&0.821384&94.7625&0.311375&160.751&0.0170099&219.575&0.608268&9385.54 + nd lab - ppo&0.80559&98.7814&0.325456&157.743&0.0175545&219.332&7.53674@xmath38@xmath39&7301.7 + ej-301&0.715235&1.29988&0.464991&203.739 & 0.0302529&0.484641&1.99773&14719.6 + @ccccc liquid&@xmath40&@xmath41&@xmath42&@xmath43 + lab - ppo&-209.56&2.81043&-0.0124283&@xmath44 + nd lab - ppo&-206.13&2.76412&-0.0122203&@xmath45 + ) in the range 2301000 nm and to a 3@xmath46 order polynomial in the range 210230 nm .
( a ) : lab - ppo .
( b ) : nd - doped lab - ppo.,title="fig:",width=283 ] ) in the range 2301000 nm and to a 3@xmath46 order polynomial in the range 210230 nm .
( a ) : lab - ppo .
( b ) : nd - doped lab - ppo.,title="fig:",width=283 ] the ej-301 results are shown in figure [ figure : ne213results](a ) .
there appears to be a region of anomalous dispersion just below 230 nm .
the measurement uncertainty increased markedly below 220 nm . to parameterize the ej-301 refractive index
, we fit to the sellmeier equation ( equation [ equation : sellmeier ] , with @xmath47=4 ) in the region 3001000 nm , and to two separate 4@xmath48 order polynomials in the regions 210240 nm and 240300 nm .
the sellmeier coefficients are given in the third row of table [ table : sellmeiercoeff ] , while the 4@xmath48 order polynomial coefficients are shown in table [ table : ne213polycoeff ] .
the fit residuals are shown in figure [ figure : ne213results](b ) .
order polynomials in the wavelength regions 210240 nm and 240300 nm , and to the sellmeier equation in the region 3001000 nm.,title="fig:",width=283 ] order polynomials in the wavelength regions 210240 nm and 240300 nm , and to the sellmeier equation in the region 3001000 nm.,title="fig:",width=283 ] @cccccc wavelength range&@xmath40&@xmath41&@xmath42&@xmath43&@xmath49 + 210240 nm&11784.7388&-205.807243&1.34648820&-3.9107264@xmath3810@xmath50&4.25436903@xmath3810@xmath51 + 240300 nm&183.668&-2.51192&0.0129879&-2.98047@xmath3810@xmath52&2.55974@xmath3810@xmath53 +
we performed ellipsometric measurements of the refractive indices of ej-301 , lab - ppo and nd - doped lab - ppo scintillators between 210 and 1000 nm .
our measurements are accurate to @xmath0 .
empirical formulae for the index of each scintillator are given .
this work was supported by the united states department of energy grant # de - fg02 - 97er41020 .
we thank our brookhaven national laboratory ( bnl ) collaborators for sending us the lab scintillator samples , and s. asahi and m. yeh for sharing the measurements performed at queen s university and bnl in the optical region .
part of this work was conducted at the university of washington nanotech user facility , a member of the nsf national nanotechnology infrastructure network ( nnin ) . | we report on ellipsometric measurements of the refractive indices of lab - ppo , nd - doped lab - ppo and ej-301 scintillators to the nearest @xmath0 , in the wavelength range @xmath1@xmath2 nm . |
a net flow of particles under unbiased external fluctuations / drive due to an asymmetric potential , which is called ratchet effect , has received much attention during the last decades .
the transport and dynamical properties of particles on asymmetric potential have been widely studied , e.g. , in physics and biology @xcite .
random motion of particles can be rectified in such an asymmetric system , which can be used for , e.g. , controlling particles motion , separating different types of particles ( i.e. , molecular sieves ) , for both underdamped and overdamped particles @xcite and for molecular motors @xcite .
vortices in a type ii superconductor often ( e.g. , for magnetic field close to @xmath3 ) can be treated as classical overdamped particles " .
most of the experiments on vortex motion rectification used arrays of asymmetric pinning sites ( e.g. , nanoengineered antidots or triangular magnetic dots / inpurties ) to create an asymmetric potential , which rectifies the motion of vortices @xcite . the rectified vortex motion was directly observed in experiments by imaging vortices via lorentz microscopy @xcite .
periodic arrangement of point defects of a gradual density or periodic square array of ferromagnetic dots of decreasing size , i.e. , varying the density of pinning sites or the size of pinning sites , were shown to result in a ratchet potential @xcite .
when vortices are trapped by pinning sites , the repulsive vortex - vortex interaction creates a higher energy barrier near the area with higher density of pinning sites .
therefore , an asymmetric potential can be created by the gradient of the density of pinning sites @xcite . even without spatial asymmetry ( i.e. , without any asymmetric walls / boundaries or asymmetric pinning sites ) ,
the motion of vortices still can be controlled by time - asymmetric driving force @xcite . due to the possibility of controlling their motion ,
the dynamical behavior of vortices in such systems has attracted considerable interest .
a series of elastic and plastic vortex flow phases were found @xcite . besides the liquid - like and solid lattice phase , vortex motion also revealed a jamming behavior @xcite .
when the density of vortices is changed , the vortex flowing direction can change to the opposite @xcite , which means vortices can drift in either the hard " direction or the easy " direction of the ratchet , depending on the vortex density . by controlling the motion of vortices ,
it is possible to remove vortices or reduce the vortex density by using a combination of two opposite oriented ratchet arrays @xcite .
the order of vortices and commensurability between vortices and cells also play an important role in vortex dynamics @xcite . in two -
dimensional ( 2d ) ratchets , the dynamics in the transverse ratchet was also studied in theory @xcite and in experiments @xcite . in the present paper ,
we study the dynamics of vortices in a circular channel formed by asymmetric triangular ( funnel ) cells ( tcs ) ( see fig .
[ fig : geo ] ) [ note that earlier this approach to form asymmetric channels in experiment ( i.e. , using weak - pinning channels ) was employed in a stripe geometry @xcite ] . due to the radially flowing current in a corbino setup ,
the driving force is not uniform inside the cell which is different from linear ratchet channels .
such a geometry ( i.e. , asymmetric in the azimuthal direction and in the radial direction ) , as will be shown , leads to a specific dynamical behavior , for example , a vortex located near the inner corner of a tc ( which is closer to the center ) , experiences a stronger driving force and moves to the next tc while a vortex located near the outer corner of tc does not move .
the circular geometry of the ratchet channel is convenient for studying commensurability and step - motor ( phase locking ) behavior @xcite .
we analyze in detail low and high density regimes of rectification , i.e. , single - vortex " and multi - vortex " regimes .
we demonstrate that the mechanism of rectification are qualitatively different for these two cases .
in addition , we have performed preliminary measurements using a single nanofabricated weak - pinning ratchet channel of a - nbge with strong - pinning nbn channel edges in a corbino set - up , and we observed a substantial asymmetric vortex response .
this paper is organized as follows .
first in sec .
ii , the model of our systems will be presented . then in sec .
iii , we will show how the density of vortices and the frequency of driving current influences the dynamical behavior of vortices .
commensurability effects of vortex density and of the frequency of current will be discussed in sec .
we present the measurements of a vortex ratchet in a corbino geometry in sec .
v. finally , the conclusions will be given in sec .
we consider a ring - like weak - pinning channel constructed of @xmath4 partially overlapping tcs as shown in fig .
[ fig : geo ] . in our simulations ,
the radius of the ring @xmath5 is typically set as @xmath6 , where @xmath7 is the magnetic field penetration depth , the wider part of the channel ( i.e. , the base of tcs ) @xmath8 , and the width of the narrow part @xmath9 ( the neck ) is typically 0.1 of the wider part ( i.e. , the ratio @xmath10 ) .
we also performed simulations in a channel with a wider neck part , e.g. , @xmath11 , for comparison .
the weak - pinning channel ( where vortices can move freely ) is surrounded by a strong - pinning superconducting material @xcite which is modeled by a medium where vortices can not move .
an external current @xmath0 radially flows from the center of the disk to the edge , resulting in the density of current @xmath12 .
therefore , the closer a vortex to the center of the disk , the stronger the lorentz force that acts on it .
the driving force due to the radial current is @xcite @xmath13 where @xmath14 is the unit vector in the azimuthal direction in the disk plane , @xmath15 is the unit of force , @xmath16 is the dimensionless driving current , and @xmath17 is the flux quantum .
we perform langevin - type molecular dynamics ( md ) simulations and numerically integrate the overdamped equations of motion @xcite : @xmath18 with @xmath19 where @xmath20 is the dimensionless viscosity coefficient which is set here to unity .
we note that in practice the viscosity coefficient @xcite varies between different superconductors . as an example , parameters for a - nbge , a typical weak - pinning material , correspond to @xmath21 ns / m@xmath22 @xcite . using this value of @xmath20 in our calculations results in typical maximum values of vortex linear velocity @xmath23
m / s ( for a 1@xmath24m - thick film ) , which approaches , but does not exceed , the larkin - ovchinnikov critical velocity for this material @xcite .
the vortex interaction @xmath25 is described by a first order of the modified bessel function of the second kind @xmath26 , and the thermal force obeys @xmath27 and @xmath28 + + to model the vortex - boundary interaction , we assume an infinite potential wall at the boundary ( i.e. , vortices can not leave the channel ) which decays inside the triangular cell with the same dependence on position as the vortex - vortex interaction potential . the total interaction of a vortex with the channel boundaries
is calculated by integrating the vortex - wall interaction potential over the geometrical boundary of the channel .
the resulting potential due to the boundary is shown in fig .
[ fig : potential ] , and the vortex - boundary force @xmath29 is directly calculated from the potential . in our simulations
, we first set @xmath30 , when no current is applied , and then gradually decrease temperature to let the system to relax to the ground state . then we set @xmath31 and apply an external driving , i.e. , an ac current resulting in oscillating lorentz force with frequency @xmath32 and amplitude @xmath33 that acts on vortices , to study the dynamics of the system .
the vortex density , i.e. , the number of vortices per tc , plays an important role in the dynamics of the system . for very low vortex density ,
the vortex - vortex interaction is small as compared to the interaction with the boundary , and therefore it can be neglected . with increasing the vortex density
, the interaction between vortices becomes important .
therefore , we roughly distinguish two regimes , i.e. , a single - vortex regime and a multi - vortex regime , respectively , corresponding to low and high vortex density . in order to characterize the dynamical behavior of the system of vortices ,
we calculate the net angular velocity of each vortex and take average over all the vortices for , e.g. , 100 ac periods .
the resulting average angular velocity ( called further `` angular velocity '' ) is denoted as @xmath1 which is analyzed for different parameters of the system and drivings .
we note that angular velocity can be related in a straightforward way to the flux - flow voltage @xcite that one would typically measure in an experiment .
one system contains @xmath34 vortices and @xmath4 triangles in a circular chain , e.g. , in our simulations we take @xmath35 . if there is less than one vortex per cell , i.e. , @xmath36 in our system ( see fig . [
fig : geo ] ) , then the density of vortices is low enough and one can neglect the interaction between vortices .
one can imagine that all the vortices are far away from each other and weakly interact with other vortices but they are strongly influenced by the ratchet potential induced by the boundary .
therefore , the dynamical behavior in the low density case is similar to that of the system with just one vortex [ shown in fig .
[ fig : iw_incom ] ( a ) ] , which is considered in the single - vortex regime .
thus , for @xmath37 , the maximum of the angular velocity is the same value [ see fig . [
fig : iw_incom ] ( a ) ] . when the density increases and therefore the interaction between the vortices becomes appreciable , the maximum of the angular velocity @xmath38 first decreases to 0.05 for @xmath39 [ see fig .
[ fig : iw_incom ] ( b ) ] and then starts to increase [ see fig .
[ fig : iw_incom ] ( c ) ] when the system turns to the regime of high vortex density . to understand this non - monotonic behavior of the maximum of the angular velocity @xmath38 [ shown in fig .
[ fig : iw_incom](d ) ] , we studied the trajectories of moving vortices .
when the system is in the single - vortex regime , where @xmath40 , all the vortices move along a nearly circular trajectory no matter whether the applied current is small or large , which can be considered as a one - dimensional ( 1d ) motion ( shown in fig .
[ fig : trace_low ] ) . for low driving currents ,
each vortex oscillates near its initial position inside a tc [ e.g. , see fig . [
fig : trace_low ] ( a ) ] , and when the driving force reaches some critical value , all the vortices move with a net angular velocity @xmath1 in the easy direction [ e.g. , see fig .
[ fig : trace_low ] ( b ) ] .
however , in the multi - vortex regime , i.e. , the high density case ( e.g. , @xmath41 ) , the motion of vortices is not 1d any more . the ac current drives vortices to pass through the narrow part ( i.e. , neck ) of the channel in the easy direction and when current is alternated , some of vortices are forced to move into the corner by the others and freeze " thus blocking the motion in the hard direction [ shown in fig .
[ fig : trace_high ] ] . due to
the asymmetry in the radial direction and radially decreasing current density in the corbino setup , a vortex near the inner corner , which is closer to the center of the disk , moves faster than the one near the outer corner .
therefore , at a specific value of the current [ e.g. , at @xmath42 , as shown in fig . [
fig : trace_high](a ) ] , the vortex near the inner corner has a larger angular velocity . when the current increases , more and more vortices move along the circular trajectory and the distribution of vortices becomes strongly inhomogeneous [ shown in fig .
[ fig : trace_high](b ) ] . with further increasing current
, this 2d trajectory becomes narrow and all the vortices move along the circle , i.e. , the 2d motion becomes 1d [ shown in fig .
[ fig : trace_high](c ) ] . with increasing density ,
more and more vortices remain in the area near the corners of tcs and block other vortices moving in the hard direction . therefore , the maximum of vortex angular velocity increases . even in the descending - velocity region of the single - vortex regime
[ see e.g. , when @xmath43 in fig .
[ fig : iw_incom](c ) ] , the angular velocity is still large .
the maximum of the angular velocity @xmath38 for different vorticities @xmath34 is shown in fig .
[ fig : iw_incom](d ) . in the single - vortex regime , where there is less than one vortex per tc
, @xmath38 remains almost the same value because of the weak interaction between vortices .
when @xmath34 becomes larger , there are more than one vortex per tc and the repulsive vortex - vortex interaction in the same tc makes the motion in either easy or hard directions much easier . therefore , @xmath38 first decreases when @xmath44 .
however , for @xmath45 , the maximum @xmath38 starts to increase .
the reason for this behavior is the transition from 1d motion to 2d motion . when vortices move along 2d trajectories ( see fig .
[ fig : iw_incom ] ) , the repulsive interaction between vortices still helps " a vortex to move to the next tc in the easy direction but blocks the vortex motion in the hard direction until applying a large enough driving current .
the radial asymmetry facilitates the motion of vortices from inner / outer corners ( driven by stronger / weaker force ) one by one .
this is different from the case of a linear ratchet channel where two vortices in the corners are driven by the same force and , as a result , arrive simultaneously at the neck region leading to jamming . in an asymmetric channel , the symmetry is broken , and jamming occurs only for rather large driving force .
the dynamical behavior of vortices in the considered ratchet system is also strongly influenced by frequency of the ac current . in this subsection
we study frequency dependence of the rectified vortex motion . for different frequencies @xmath32 ,
the @xmath46 curves are shown in fig .
[ fig : iw_gap ] .
when the frequency is low , the dynamical behavior is similar for different vortex densities . as the drive amplitude @xmath47
is increased , the angular velocity first increases , reaches a maximum , and then decreases to zero .
the curves , especially for @xmath48 , in fig .
[ fig : iw_gap](a ) are similar to the analytic result in a ratchet potential ( e.g. , see fig . 2 in ref .
@xcite ) . in the case of @xmath48 , the first critical value of current @xmath49 [ shown in fig .
[ fig : iw_gap](a ) ] corresponds to the maximum friction in the easy direction @xmath50 and the second one @xmath51 corresponds to the case when the driving force reaches the maximum friction in the hard direction @xmath52 .
when @xmath53 , the first critical value of current decreases to @xmath54 , which means vortices are easier to move in the easy direction
. if the density increases further , @xmath54 decreases and @xmath55 increases , for @xmath41 [ shown in fig .
[ fig : iw_gap](a ) ]
. therefore , the interaction between vortices allows vortices to move even easier in the easy direction and harder in the hard direction . for an intermediate frequency [ e.g. , @xmath56 as shown in fig .
[ fig : iw_gap](c ) ] , when the distance a vortex moves during one period is comparable to the size of a tc , the dynamical behavior of vortices will depend not only on the driving force but also on the vortex density .
this is explained by the fact that the confinement force due to the boundary and the vortex - vortex interaction are also comparable . in this case
we obtain different dynamical behavior in single - vortex and multi - vortex regimes which were discussed in the previous subsection . due to the commensurability between the numbers of tcs and vortices , a step structure of the @xmath46 curve
is revealed in the low density case ( e.g. , @xmath48 and @xmath57 ) , which will be discussed in more detail in sec .
as one can predict , the angular velocity becomes zero [ see fig . [
fig : iw_gap](e ) ] when the frequency is high enough because a vortex oscillates near its initial position inside a cell and thus its motion can not be rectified .
therefore , the variation of the angular velocity is much smaller than that in the low / intermediate frequency case .
each vortex is localized in a specific tc , i.e. , the vortex is only influenced by a single potential well but not by the ratchet potential . for comparison
, we also calculate @xmath46 curves for the case when a tc has a wider connection part ( @xmath11 ) [ shown in fig .
[ fig : iw_gap](b ) , ( d ) , ( f ) ] .
the function @xmath58 in general shows a similar behavior as for @xmath59 .
however , for the intermediate frequency ( @xmath56 ) of the applied current , we obtain more steps in @xmath46 curve in the case of @xmath48 [ shown in fig .
[ fig : iw_gap](d ) ] than that for @xmath59 [ shown in fig .
[ fig : iw_gap](c ) ] .
this relates to the commensurability effect that will also be discussed in sec .
as we defined above , the system contains @xmath34 vortices and @xmath4 triangles in a circular chain . if there is a common integer ( except one ) in terms of which
two numbers @xmath34 and @xmath4 can both be measured , then they are commensurate .
otherwise , they are incommensurate .
[ fig : iw_com ] ) shows the average angular velocity @xmath1 as a function of @xmath33 for different commensurate ratios .
the ascending part of the @xmath58 curves ( i.e. , where @xmath1 versus @xmath33 increases ) is stepwise .
besides the large steps of angular velocity in the @xmath46 curve for @xmath48 , which we refer to as integer steps " , we also found smaller steps for some specific vorticities @xmath34 ( shown in fig .
[ fig : iw_com ] ) . if @xmath60 and @xmath61 , where @xmath62 and @xmath63 are incommensurate integers , the small steps can be found in the @xmath46 curves [ in fig .
[ fig : iw_com](b)-(d ) ] .
when @xmath64 , we observe only integer steps of the angular velocity @xmath58 [ in fig .
[ fig : iw_com](a ) ] .
the difference in the angular velocity between two adjacent steps is always @xmath65 .
if @xmath61 , then we can find a fractional step whose magnitude is @xmath66 .
for example , for @xmath67 [ shown in fig .
[ fig : iw_com](b ) ] , the smallest step is @xmath68 in the system with 72 vortices ( i.e. , @xmath69 ) , which is a half of that for @xmath64 ( e.g. , when @xmath70 ) . in the case of @xmath64 ,
a unit cell contains only one vortex .
[ the unit cell " ( uc ) is a minimum repeatable set of tc(s ) containing an integer number of vortices .
for example , for @xmath48 the uc is the entire channel with one vortex , and for @xmath70 the uc is one tc with one vortex . ]
if one vortex can overcome the potential barrier and move in the easy direction , all the vortices can do so at the same time , i.e. , collectively .
this results in integer steps of the average angular velocity @xmath65 in the @xmath46 curve .
however , fractional steps appear when there are more than one vortex in each uc . for instance , a uc contains one tc with two vortices when @xmath69 , and if @xmath71 , the uc is constructed by three tcs with two vortices . in general case
, the unit cell contains @xmath63 tcs with @xmath62 vortices inside .
if @xmath72 , there are more than one vortex in the uc , and those vortices are not equivalent , i.e. , they are not located at the equivalent position in the tc and/or they experience different interactions with the boundary .
therefore , they move with different angular velocities in each period .
let us take @xmath67 , for example ( see fig .
[ fig : typeab ] ) . as shown in fig .
[ fig : typeab](a ) , the vortices located in a tc that is followed by an empty tc in the easy direction ( type - a ) can overcome the potential barriers in a period of alternating current and move to its neighbor tc , while the others ( type - b ) are still localized in its original tc . when @xmath73 [ shown in fig .
[ fig : typeab](b ) ] , type - a vortices move from c2 to c3 but type - b vortices only oscillate in c1 . in the view of the whole circular channel ,
only a half of vortices ( i.e. , type - a vortices ) move to another tc while the other half of vortices ( i.e. , type - b vortices ) do not move .
after that , the type - b vortices are located at similar position as type - a vortices in the previous cycle , i.e. , the previous type - b vortices now become type - a vortices and they will move to next tc in the anticlockwise direction during the next period . therefore , when we calculate the average angular velocity for all the vortices , the value is a half of the angular velocity of type - a vortices , i.e. , @xmath68 .
if the current increases , every vortex can overcome the barriers and move to the neighbor tc [ e.g. , as shown in fig .
[ fig : typeab ] , type - a vortices move from c2 to c3 and type - b vortices move from c1 to c2 when @xmath74 . then average angular velocity becomes @xmath65 which is the same as that for @xmath64 when all vortices move to the neighbor tc .
curve for a channel with @xmath75 and different @xmath62 : @xmath76 1 ( a ) , 2 ( b ) , 3 ( c ) , and 5 ( d ) .
@xmath34 is the vorticity , i.e. , the number of vortices .
@xmath4 is the number of cells in the circular chain .
if @xmath77 , e.g. , @xmath78 , the @xmath46 curve shows several steps and the height of each integer step is @xmath65 . when @xmath60 ( k=2 , 3 , 5 ) , the height of each fractional step is @xmath66 [ e.g. , as shown in ( b ) ( c ) and ( d)].,scaledwidth=48.0% ] further we analyze the angular velocity evolution while varying the ac drive frequency ( shown in fig .
[ fig : fw ] ) . for varying frequency @xmath32 ,
the @xmath79 curves are characterized by peaks and/or oscillations .
let us clarify this behavior . as discussed in sec .
iii , if @xmath80 , vortices can move in the easy direction and be frozen / blocked in the hard direction when the current alternates .
let us introduce a time scale , @xmath81 , to characterize the motion of a vortex over the entire tc .
one tc occupies an angle @xmath82 .
if we assume that a vortex moves over the entire tc to the equivalent position in the next tc , then @xmath81 should satisfy the following condition : @xmath83\cdot\hat{\theta}dt/ \eta r= \theta_0,\ ] ] where @xmath14 is the unit vector in the azimuthal direction .
in the single vortex case ( @xmath48 ) , we set @xmath84 , and @xmath85 is a constant . considering the circular trajectory of the moving vortex ,
the integral of @xmath86 is also a constant .
therefore , @xmath81 only depends on the current @xmath47 and we use the notation @xmath87 instead of @xmath81 in order to show the dependence on the driving current .
if @xmath88 and the ac period @xmath89 , then the rectified signal will be maximum since all the vortices coherently move over the entire tc in the easy direction during the first half period and do not move backwards in the next half period . then we can obtain the principal period @xmath90 for each value of the current and roughly estimate forces @xmath91 and @xmath92 from the @xmath79 or @xmath93 curves [ as shown in fig .
[ fig : fw ] ] .
the maximum angular velocity @xmath38 for different current and vorticities @xmath34 is shown in the insets of fig .
[ fig : fw ] . for example
, in the case of @xmath48 [ see fig .
[ fig : fw](a ) ] , the angular velocity is always zero when @xmath94 and becomes non - zero for @xmath73 , i.e. , the vortex starts to move in the easy direction . therefore , @xmath95 .
however , for @xmath96 , the maximum angular velocity @xmath38 decreases as compared to the case for @xmath97 , which means @xmath98 . for @xmath99
when driving current increases , vortices move backwards ( i.e. , the motion in the hard direction ) therefore resulting in the decreasing net angular velocity .
for example , when @xmath100 [ shown in fig .
[ fig : fw](a ) ] , the maximum @xmath38 becomes smaller and the jumps in the @xmath93 curves become smoother , which means the effect of boundary becomes weaker under a strong driving force .
therefore , the angular velocity should be zero when the driving force goes to infinity .
it explains why the angular velocity decreases and goes to zero when the driving current increases above some critical value . for the single - vortex case ,
the frequency dependence of the first local maximum @xmath101 in the @xmath102 curves versus the corresponding drive frequencies , @xmath103 , is plotted in fig .
[ fig : line ] which shows a linear behavior . when the density is increased up to one vortex per cell ( @xmath104 )
, the @xmath93 curve is similar to the one for @xmath48 for both incommensurate [ e.g. , see fig .
[ fig : fw](b ) ] and commensurate cases [ see fig . [ fig : fw](c ) ] because the interaction between vortices is weak . for a higher density ,
the ( in-)commensurability effect becomes pronounced .
the sharp jumps of the angular velocity are obtained only in the commensurate case [ e.g. , @xmath69 shown in fig .
[ fig : fw](e ) ] but not in the incommensurate case [ e.g. , @xmath41 shown in fig .
[ fig : fw](f ) ] . comparing the high density cases with the low density cases
, we can conclude that sharp jumps in @xmath93 curve , which are both found in single - vortex and multi - vortex regimes , are due to the effect of periodically repeated boundaries of the ratchet potential and the increasing local maximum of the angular velocity when @xmath105 increases in the multi - vortex regime .
this is different from the single - vortex regime , because of the strong interaction between vortices induced by the high vortex density .
if the principal period ( i.e. , the time period for the first jump ) is denoted by @xmath106 , the others will be harmonics of @xmath106 , @xmath107 . of the @xmath93 curves for vortices in the single - vortex regime when the driving force is smaller than @xmath108 .
the magnitude of the first peak shows a linear dependence on the principal frequency @xmath109.,title="fig:",scaledwidth=45.0% ] +
we have performed preliminary measurements on a device containing a single weak - pinning circular channel for guiding vortex motion in a corbino geometry .
this is based on a technique for using weak - pinning channels with tailored edges to produce asymmetric vortex confining potentials .
such an arrangement resulted in substantial asymmetric vortex response for linear channels on a strip geometry @xcite .
the corbino sample consists of a si substrate with a 200-nm thick film of weak - pinning a - nbge and a 50-nm thick film of strong - pinning nbn on top .
the fabrication followed the scheme of previous weak - pinning channel devices @xcite , with the 1.5-mm diameter corbino disk pattern etched through the entire superconducting bilayer .
the 500-@xmath24 m diameter circular chain of triangular cells was etched through the nbn layer ( fig .
[ fig : exp01 ] ) , thus defining the weak - pinning a - nbge channel region for vortex flow .
+ wirebonds were attached between the center and perimeter of the corbino disk for injecting a bias current with a radial flow . because of the rather small flux - flow voltages for vortex motion in a single channel , it was necessary to use a custom picovoltmeter based on a dc squid operated in a flux - locked loop @xcite .
measurements of the noise power at different temperatures were used to calibrate the value for the series resistance at the squid input , thus allowing for a measurement of the system gain , as described in ref .
@xcite . during the measurements ,
the sample and squid were immersed in a pumped liquid helium bath .
shielding of external magnetic fields was achieved with a @xmath24-metal shield surrounding the dewar and a superconducting pb shield around the sample and squid on the bottom of the cryogenic insert .
vortices were introduced into the channel by temporarily raising the sample above the helium bath , heating to @xmath1106 k between @xmath111 = 2.88 k and @xmath112 @xmath113 10 k while applying a small magnetic field with a helmholtz coil on the insert , then cooling back down below @xmath111 .
+ upon reaching the desired measurement temperature , the bias current was varied incrementally and the flux - flow voltage sensed by the squid was recorded for each current value .
such current - voltage characteristics ( ivc ) exhibited substantial asymmetries between the critical current for vortices to begin to move through the channel for the two directions .
this asymmetry persisted down to the lowest measurement temperature , 1.60 k , well below the onset of superconductivity in the channel at @xmath111 ( fig .
[ fig : exp02 ] ) .
the larger critical current corresponded to the sense of vortex motion in the hard direction through the ratchet cells . due to experimental wiring limitations on these preliminary measurements of a corbino ratchet channel
, it was not possible to sweep the bias current with an oscillatory drive to study the average flux - flow voltage over a cycle .
nonetheless , the large asymmetry between the two senses of critical current demonstrates the potential for a weak - pinning ratchet channel in a corbino geometry to rectify vortex motion .
a vortex moving in an asymmetric circular channel in a corbino setup experiences the confinement of the boundary , the repulsive interaction due to other vortices and the gradient ( i.e. , radially decreasing ) driving force when an external current is applied .
the combination of these factors determines the vortex motion .
different dynamical behavior is observed in low and high density systems , which are referred to as single - vortex " and multi - vortex " regimes , respectively . for low density ,
i.e. , in the single - vortex regime , there is no more than one vortex per triangular cell so that the vortex - vortex interaction is negligible
. therefore , the ratchet potential due to the boundary dominates and all the vortices follow circular 1d trajectories . since all the vortices still move in 1d when there is more than one vortex per cell ( i.e. , a higher density ) , a vortex can escape even easier either in the easy or in the hard direction due to the interaction between vortices in the same cell .
this results in decrease of the rectified net current .
however , when the number of vortices increases further ( i.e. , in themulti - vortex " regime of rectification when the rectified net current increases with the vortex density ) , the trajectories of vortex motion for low driving currents are 2d while these trajectories squeeze and turn to 1d circle with increasing driving current .
because of the circular geometry of the channel , the density of vortices first becomes inhomogeneous [ i.e , see fig .
[ fig : trace_high](b ) ] during the transition from 2d motion to 1d motion in the multi - vortex regime and then becomes homogeneous again when all the vortices move in a circular trajectory .
considering the asymmetry in the radial direction , vortices near the inner / outer corner of the triangular cells ( tcs ) ( i.e. , closer / further to the center of the disk ) are driven by different lorentz forces and for some specific value of driving current the vortex in the outer corner moves to the inner corner while the one in the inner corner moves to the next tc [ e.g. , see fig .
[ fig : trace_high](a ) ] .
this kind of motion prevents vortices from arriving simultaneously at the narrow part that would lead to jamming which occurs in case of a linear channel . when the density increases , the maximum net angular velocity @xmath38 remains the same in the single - vortex regime and then decreases until reaching the multi - vortex regime , and then @xmath38 increases .
the frequency of driving current also strongly influences the vortex dynamical behavior .
the ac frequency determines the possible distance a vortex moves during an ac period . for high frequency ,
a nearly zero net angular velocity is obtained for different values of driving current @xmath47 .
each vortex is unable to overcome the energy barrier and is confined by a single potential well instead of the periodic ratchet potential .
when the frequency is low , the ratchet effect is clearly observed in the @xmath46 curve but the commensurability effect is not present . for an intermediate frequency of driving current
under which the distance of moving vortex in a period is comparable to the size of the tc , both the ratchet effect and the commensurability effect have been observed . besides the vortex density and the frequency of current , the commensurability between the numbers of vortices and tcs also plays an important role in the dynamical behavior of vortices , which leads to jumps in the angular velocity @xmath1 with increasing driving current @xmath47 ( i.e. , steps in the @xmath46 curve ) .
therefore , under some specific conditions , the average angular velocity of vortices is not a continuous function of the driving current .
the commensurability also influences the minimum difference of the angular velocity for different steps in the @xmath46 curve , and results in integer steps ( i.e. , the large steps which are found , e.g. , for vorticity @xmath48 ) and fractional steps ( i.e. , the smaller steps whose magnitude are fractions of the magnitude of the integer steps , e.g. , for vorticity @xmath71 ) in a certain range of the current .
we also obtained several peaks ( sharp jumps ) in the @xmath93 curve , which correspond to the principal period @xmath106 ( during which the vortex can move over one tc in the easy direction ) , and harmonics periods @xmath114 ( during which vortices move over @xmath62 tcs in the easy direction ) .
the net flow of vortices is enhanced when the ac period is one of the harmonics periods , i.e. , the average angular velocity reaches a local maximum .
we thank peter kes and marcel hesselberth for providing the superconducting films from which the corbino ratchet sample was fabricated . this work was supported by the `` odysseus '' program of the flemish government and the flemish science foundation ( fwo - vl ) , the interuniversity attraction poles ( iap ) programme belgian state belgian science policy , and the fwo - vl ( belgium ) .
, k.y . , and b.l.t.p acknowledge support from the national science foundation under grant dmr-0547147 , as well as the use of the cornell nanoscale facility , a member of the national nanotechnology infrastructure network , which is supported by the national science foundation ( grant ecs-0335765 ) . | we study the dynamics of vortices in an asymmetric ( i.e. , consisting of triangular cells ) ring channel driven by an external current @xmath0 in a corbino setup .
the asymmetric potential can rectify the motion of vortices and cause a net flow without any unbiased external drive , which is called ratchet effect . with an applied ac current , the potential can rectify the motion of vortices in the channel and induce a dc net flow .
we show that the net flow of vortices in such a system strongly depends on vortex density and frequency of the driving force / current .
depending on the density , we distinguish a single - vortex " rectification regime ( for low density , when each vortex is rectified individually ) determined by the potential - energy landscape inside each cell of the channel ( i.e. , hard " and easy " directions of motion ) and multi - vortex " , or
collective " , rectification ( high density case ) when the interaction between vortices becomes important .
the frequency of the driving ac current determines a possible distance that a vortex could move during one period . for high frequency current
, vortices only oscillate in the triangular cell .
for low frequency , the vortex angular velocity @xmath1 increases nearly linearly until the driving force reaches the maximum friction force in the hard direction .
furthermore , the commensurability between the number of vortices and the number of cells results in a stepwise @xmath2 curve , which means that the average angular velocity @xmath1 of the vortices is a discontinuous function of the applied current .
besides the integer " steps , i.e. , the large steps found in the single vortex case , we also found fractional steps corresponding to fractional ratio between the numbers of vortices and triangular cells .
the principal and fractional frequencies for different currents are found , when the net flow of vortices reaches the maximum that is proportional to the frequency when the density of vortices is low .
we have performed preliminary measurements on a device containing a single weak - pinning circular ratchet channel in a corbino geometry and observed a substantial asymmetric vortex response . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.